3 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
4 for currently supported client and server kernel versions.
5 * Server primary kernels built and tested during release cycle:
6 3.10.0-957.27.2.el7 (RHEL7.6)
7 3.10.0-1062.18.1.el7 (RHEL7.7)
8 3.10.0-1127.19.1.el7 (RHEL7.8)
9 3.10.0-1160.25.1.el7 (RHEL7.9)
10 * Server known to build on the following kernels (others may also work):
11 2.6.32-696.18.7.el6 (RHEL6.9)
12 3.10.0-862.14.4.el7 (RHEL7.5)
13 4.4.120-92.70 (SLES12 SP2)
14 4.4.180-94.97 (SLES12 SP3)
15 4.4.0-142 (Ubuntu 16.04)
16 4.15.0-45 (Ubuntu 18.04)
17 vanilla linux 4.6.7 (ZFS only)
18 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
19 * Client primary unpatched kernels built and tested during release cycle:
20 3.10.0-957.27.2.el7 (RHEL7.6)
21 3.10.0-1062.18.1.el7 (RHEL7.7)
22 3.10.0-1127.19.1.el7 (RHEL7.8)
23 3.10.0-1160.25.1.el7 (RHEL7.9)
24 4.18.0-80.11.2.el8 (RHEL8.0)
25 4.4.180-94.97 (SLES12 SP3)
26 4.12.14-95.54 (SLES12 SP4)
27 4.12.14-122.66 (SLES12 SP5)
28 4.15.0-45 (Ubuntu 18.04)
29 * Client known to build on unpatched kernels (others may also work):
30 2.6.32-696.18.7.el6 (RHEL6.9)
31 2.6.32-754.6.3.el6 (RHEL6.10)
32 3.10.0-862.14.4.el7 (RHEL7.5)
33 4.14.0-49.13.1.el7a (RHEL7.5)
34 4.14.0-115.2.2.el7a (RHEL7.6)
35 4.18.0-147.3.1.el8 (RHEL8.1)
36 4.18.0-193.28.1.el8 (RHEL8.2)
37 4.18.0-240.22.1.el8 (RHEL8.3)
38 4.18.0-305.3.1.el8 (RHEL8.4)
39 4.4.120-92.70 (SLES12 SP2)
40 4.4.0-142 (Ubuntu 16.04)
41 vanilla linux 4.15.0 (ZFS only)
42 * Recommended e2fsprogs version: 1.45.6.wc1 or newer
43 * Recommended ZFS / SPL version: 0.7.13
44 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
45 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
46 stack overflow. For more information, please refer to bugzilla 17630.
47 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
48 the re-exporting nodes to be mounted with "32bitapi" mount option
49 * Note for Ubuntu 14.04.5:
50 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
51 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
55 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
56 for currently supported client and server kernel versions.
57 * Server primary kernels built and tested during release cycle:
58 3.10.0-957.27.2.el7 (RHEL7.6)
59 3.10.0-1062.18.1.el7 (RHEL7.7)
60 3.10.0-1127.19.1.el7 (RHEL7.8)
61 3.10.0-1160.15.2.el7 (RHEL7.9)
62 * Server known to build on the following kernels (others may also work):
63 2.6.32-696.18.7.el6 (RHEL6.9)
64 3.10.0-862.14.4.el7 (RHEL7.5)
65 4.4.120-92.70 (SLES12 SP2)
66 4.4.180-94.97 (SLES12 SP3)
67 4.4.0-142 (Ubuntu 16.04)
68 4.15.0-45 (Ubuntu 18.04)
69 vanilla linux 4.6.7 (ZFS only)
70 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
71 * Client primary unpatched kernels built and tested during release cycle:
72 3.10.0-957.27.2.el7 (RHEL7.6)
73 3.10.0-1062.18.1.el7 (RHEL7.7)
74 3.10.0-1127.19.1.el7 (RHEL7.8)
75 3.10.0-1160.15.2.el7 (RHEL7.9)
76 4.18.0-80.11.2.el8 (RHEL8.0)
77 4.4.180-94.97 (SLES12 SP3)
78 4.12.14-95.54 (SLES12 SP4)
79 4.12.14-122.57 (SLES12 SP5)
80 4.15.0-45 (Ubuntu 18.04)
81 * Client known to build on unpatched kernels (others may also work):
82 2.6.32-696.18.7.el6 (RHEL6.9)
83 2.6.32-754.6.3.el6 (RHEL6.10)
84 3.10.0-862.14.4.el7 (RHEL7.5)
85 4.14.0-49.13.1.el7a (RHEL7.5)
86 4.14.0-115.2.2.el7a (RHEL7.6)
87 4.18.0-147.3.1.el8 (RHEL8.1)
88 4.18.0-193.28.1.el8 (RHEL8.2)
89 4.18.0-240.1.1.el8 (RHEL8.3)
90 4.4.120-92.70 (SLES12 SP2)
91 4.4.0-142 (Ubuntu 16.04)
92 vanilla linux 4.15.0 (ZFS only)
93 * Recommended e2fsprogs version: 1.45.6.wc1 or newer
94 * Recommended ZFS / SPL version: 0.7.13
95 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
96 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
97 stack overflow. For more information, please refer to bugzilla 17630.
98 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
99 the re-exporting nodes to be mounted with "32bitapi" mount option
100 * Note for Ubuntu 14.04.5:
101 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
102 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
106 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
107 for currently supported client and server kernel versions.
108 * Server primary kernels built and tested during release cycle:
109 3.10.0-957.27.2.el7 (RHEL7.6)
110 3.10.0-1062.18.1.el7 (RHEL7.7)
111 3.10.0-1127.8.2.el7 (RHEL7.8)
112 * Server known to build on the following kernels (others may also work):
113 2.6.32-696.18.7.el6 (RHEL6.9)
114 3.10.0-862.14.4.el7 (RHEL7.5)
115 4.4.120-92.70 (SLES12 SP2)
116 4.4.180-94.97 (SLES12 SP3)
117 4.4.0-142 (Ubuntu 16.04)
118 4.15.0-45 (Ubuntu 18.04)
119 vanilla linux 4.6.7 (ZFS only)
120 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
121 * Client primary unpatched kernels built and tested during release cycle:
122 3.10.0-957.27.2.el7 (RHEL7.6)
123 3.10.0-1062.18.1.el7 (RHEL7.7)
124 3.10.0-1127.8.2.el7 (RHEL7.8)
125 4.18.0-80.11.2.el8 (RHEL8.0)
126 4.4.180-94.97 (SLES12 SP3)
127 4.12.14-95.45 (SLES12 SP4)
128 4.12.14-122.20 (SLES12 SP5)
129 4.15.0-45 (Ubuntu 18.04)
130 * Client known to build on unpatched kernels (others may also work):
131 2.6.32-696.18.7.el6 (RHEL6.9)
132 2.6.32-754.6.3.el6 (RHEL6.10)
133 3.10.0-862.14.4.el7 (RHEL7.5)
134 4.14.0-49.13.1.el7a (RHEL7.5)
135 4.14.0-115.2.2.el7a (RHEL7.6)
136 4.18.0-147.8.1.el8 (RHEL8.1)
137 4.18.0-193.1.2.el8 (RHEL8.2)
138 4.4.120-92.70 (SLES12 SP2)
139 4.4.0-142 (Ubuntu 16.04)
140 vanilla linux 4.15.0 (ZFS only)
141 * Recommended e2fsprogs version: 1.45.2.wc1 or newer
142 * Recommended ZFS / SPL version: 0.7.13
143 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
144 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
145 stack overflow. For more information, please refer to bugzilla 17630.
146 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
147 the re-exporting nodes to be mounted with "32bitapi" mount option
148 * Note for Ubuntu 14.04.5:
149 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
150 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
154 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
155 for currently supported client and server kernel versions.
156 * Server primary kernels built and tested during release cycle:
157 3.10.0-957.27.2.el7 (RHEL7.6)
158 3.10.0-1062.9.1.el7 (RHEL7.7)
159 * Server known to build on the following kernels (others may also work):
160 2.6.32-696.18.7.el6 (RHEL6.9)
161 3.10.0-862.14.4.el7 (RHEL7.5)
162 4.4.120-92.70 (SLES12 SP2)
163 4.4.180-94.97 (SLES12 SP3)
164 4.4.0-142 (Ubuntu 16.04)
165 4.15.0-45 (Ubuntu 18.04)
166 vanilla linux 4.6.7 (ZFS only)
167 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
168 * Client primary unpatched kernels built and tested during release cycle:
169 3.10.0-957.27.2.el7 (RHEL7.6)
170 3.10.0-1062.9.1.el7 (RHEL7.7)
171 4.18.0-80.11.2.el8 (RHEL8.0)
172 4.4.180-94.97 (SLES12 SP3)
173 4.12.14-95.29 (SLES12 SP4)
174 4.15.0-45 (Ubuntu 18.04)
175 * Client known to build on unpatched kernels (others may also work):
176 2.6.32-696.18.7.el6 (RHEL6.9)
177 2.6.32-754.6.3.el6 (RHEL6.10)
178 3.10.0-862.14.4.el7 (RHEL7.5)
179 4.14.0-49.13.1.el7a (RHEL7.5)
180 4.14.0-115.2.2.el7a (RHEL7.6)
181 4.18.0-147.3.1.el8 (RHEL8.1)
182 4.4.120-92.70 (SLES12 SP2)
183 4.4.0-142 (Ubuntu 16.04)
184 vanilla linux 4.15.0 (ZFS only)
185 * Recommended e2fsprogs version: 1.45.2.wc1 or newer
186 * Recommended ZFS / SPL version: 0.7.13
187 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
188 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
189 stack overflow. For more information, please refer to bugzilla 17630.
190 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
191 the re-exporting nodes to be mounted with "32bitapi" mount option
192 * Note for Ubuntu 14.04.5:
193 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
194 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
198 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
199 for currently supported client and server kernel versions.
200 * Server primary kernels built and tested during release cycle:
201 3.10.0-957.27.2.el7 (RHEL7.6)
202 3.10.0-1062.1.1.el7 (RHEL7.7)
203 * Server known to build on the following kernels (others may also work):
204 2.6.32-696.18.7.el6 (RHEL6.9)
205 3.10.0-862.14.4.el7 (RHEL7.5)
206 4.4.120-92.70 (SLES12 SP2)
207 4.4.180-94.97 (SLES12 SP3)
208 4.4.0-142 (Ubuntu 16.04)
209 4.15.0-45 (Ubuntu 18.04)
210 vanilla linux 4.6.7 (ZFS only)
211 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
212 * Client primary unpatched kernels built and tested during release cycle:
213 3.10.0-957.27.2.el7 (RHEL7.6)
214 3.10.0-1062.1.1.el7 (RHEL7.7)
215 4.18.0-80.7.1.el8 (RHEL8.0)
216 4.4.180-94.97 (SLES12 SP3)
217 4.12.14-95.29 (SLES12 SP4)
218 4.15.0-45 (Ubuntu 18.04)
219 * Client known to build on unpatched kernels (others may also work):
220 2.6.32-696.18.7.el6 (RHEL6.9)
221 2.6.32-754.6.3.el6 (RHEL6.10)
222 3.10.0-862.14.4.el7 (RHEL7.5)
223 4.14.0-49.13.1.el7a (RHEL7.5)
224 4.14.0-115.2.2.el7a (RHEL7.6)
225 4.4.120-92.70 (SLES12 SP2)
226 4.4.0-142 (Ubuntu 16.04)
227 vanilla linux 4.15.0 (ZFS only)
228 * Recommended e2fsprogs version: 1.45.2.wc1 or newer
229 * Recommended ZFS / SPL version: 0.7.13
230 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
231 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
232 stack overflow. For more information, please refer to bugzilla 17630.
233 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
234 the re-exporting nodes to be mounted with "32bitapi" mount option
235 * Note for Ubuntu 14.04.5:
236 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
237 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
241 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
242 for currently supported client and server kernel versions.
243 * Server known to build on the following kernels (others may also work):
244 2.6.32-696.18.7.el6 (RHEL6.9)
245 3.10.0-862.14.4.el7 (RHEL7.5)
246 3.10.0-957.10.1.el7 (RHEL7.6)
247 4.4.120-92.70 (SLES12 SP2)
248 4.4.162-94.69 (SLES12 SP3)
249 4.4.0-142 (Ubuntu 16.04)
250 4.15.0-45 (Ubuntu 18.04)
251 vanilla linux 4.6.7 (ZFS only)
252 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
253 * Client known to build on unpatched kernels (others may also work):
254 2.6.32-696.18.7.el6 (RHEL6.9)
255 2.6.32-754.6.3.el6 (RHEL6.10)
256 3.10.0-862.14.4.el7 (RHEL7.5)
257 3.10.0-957.10.1.el7 (RHEL7.6)
258 4.14.0-49.13.1.el7a (RHEL7.5)
259 4.14.0-115.2.2.el7a (RHEL7.6)
260 4.4.120-92.70 (SLES12 SP2)
261 4.4.162-94.69 (SLES12 SP3)
262 4.12.14-95.13 (SLES12 SP4)
263 4.4.0-142 (Ubuntu 16.04)
264 4.15.0-45 (Ubuntu 18.04)
265 vanilla linux 4.15.0 (ZFS only)
266 * Recommended e2fsprogs version: 1.44.5.wc1 or newer
267 * Recommended ZFS / SPL version: 0.7.13
268 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
269 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
270 stack overflow. For more information, please refer to bugzilla 17630.
271 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
272 the re-exporting nodes to be mounted with "32bitapi" mount option
273 * Note for Ubuntu 14.04.5:
274 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
275 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
280 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
281 for currently supported client and server kernel versions.
282 * Server known to build on the following kernels (others may also work):
283 2.6.32-696.18.7.el6 (RHEL6.9)
284 3.10.0-862.14.4.el7 (RHEL7.5)
285 3.10.0-957.10.1.el7 (RHEL7.6)
286 4.4.120-92.70 (SLES12 SP2)
287 4.4.162-94.69 (SLES12 SP3)
288 4.4.0-142 (Ubuntu 16.04)
289 4.15.0-45 (Ubuntu 18.04)
290 vanilla linux 4.6.7 (ZFS only)
291 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
292 * Client known to build on unpatched kernels (others may also work):
293 2.6.32-696.18.7.el6 (RHEL6.9)
294 2.6.32-754.6.3.el6 (RHEL6.10)
295 3.10.0-862.14.4.el7 (RHEL7.5)
296 3.10.0-957.10.1.el7 (RHEL7.6)
297 4.14.0-49.13.1.el7a (RHEL7.5)
298 4.14.0-115.2.2.el7a (RHEL7.6)
299 4.4.120-92.70 (SLES12 SP2)
300 4.4.162-94.69 (SLES12 SP3)
301 4.12.14-95.13 (SLES12 SP4)
302 4.4.0-142 (Ubuntu 16.04)
303 4.15.0-45 (Ubuntu 18.04)
304 vanilla linux 4.15.0 (ZFS only)
305 * Recommended e2fsprogs version: 1.44.5.wc1 or newer
306 * Recommended ZFS / SPL version: 0.7.13
307 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
308 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
309 stack overflow. For more information, please refer to bugzilla 17630.
310 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
311 the re-exporting nodes to be mounted with "32bitapi" mount option
312 * Note for Ubuntu 14.04.5:
313 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
314 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
318 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
319 for currently supported client and server kernel versions.
320 * Server known to build on the following kernels (others may also work):
321 2.6.32-696.18.7.el6 (RHEL6.9)
322 3.10.0-862.14.4.el7 (RHEL7.5)
323 3.10.0-957.el7 (RHEL7.6)
324 4.4.120-92.70 (SLES12 SP2)
325 4.4.162-94.69 (SLES12 SP3)
326 4.4.0-131 (Ubuntu 16.04)
327 4.15.0-32 (Ubuntu 18.04)
328 vanilla linux 4.6.7 (ZFS only)
329 * ldiskfs needs an ldiskfs patch series for that kernel, ZFS does not
330 * Client known to build on unpatched kernels (others may also work):
331 2.6.32-696.18.7.el6 (RHEL6.9)
332 2.6.32-754.6.3.el6 (RHEL6.10)
333 3.10.0-862.14.4.el7 (RHEL7.5)
334 3.10.0-957.el7 (RHEL7.6)
335 4.14.0-49.13.1.el7a (RHEL7.5)
336 4.14.0-115.2.2.el7a (RHEL7.6)
337 4.4.120-92.70 (SLES12 SP2)
338 4.4.162-94.69 (SLES12 SP3)
339 4.4.0-131 (Ubuntu 16.04)
340 4.15.0-32 (Ubuntu 18.04)
341 vanilla linux 4.15.0 (ZFS only)
342 * Recommended e2fsprogs version: 1.44.3.wc1 or newer
343 * Recommended ZFS / SPL version: 0.7.9
344 * Tested with ZFS / SPL version: 0.7.9
345 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
346 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
347 stack overflow. For more information, please refer to bugzilla 17630.
348 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
349 the re-exporting nodes to be mounted with "32bitapi" mount option
350 * Note for Ubuntu 14.04.5:
351 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
352 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
354 --------------------------------------------------------------------------------
357 2018-04-03 Intel Corporation
359 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
360 for currently supported client and server kernel versions.
361 * Server known to build on patched kernels:
362 2.6.32-431.29.2.el6 (RHEL6.5)
363 2.6.32-504.30.3.el6 (RHEL6.6)
364 2.6.32-573.26.1.el6 (RHEL6.7)
365 2.6.32-642.15.1.el6 (RHEL6.8)
366 2.6.32-696.18.7.el6 (RHEL6.9)
367 3.10.0-693.21.1.el7 (RHEL7.4)
368 3.0.101-0.47.71 (SLES11 SP3)
369 3.0.101-107 (SLES11 SP4)
370 3.12.74-60.64.40 (SLES12 SP1)
371 4.4.114-92.64 (SLES12 SP2)
372 4.4.114-94.11 (SLES12 SP3)
373 3.13.0-101 (Ubuntu 14.04, ZFS only)
374 4.4.0-85.108 (Ubuntu 14.04.5 LTS)
375 4.4.0-116 (Ubuntu 16.04)
376 vanilla linux 4.6.7 (ZFS only)
377 * Client known to build on unpatched kernels:
378 2.6.32-431.29.2.el6 (RHEL6.5)
379 2.6.32-504.30.3.el6 (RHEL6.6)
380 2.6.32-573.26.1.el6 (RHEL6.7)
381 2.6.32-642.15.1.el6 (RHEL6.8)
382 2.6.32-696.18.7.el6 (RHEL6.9)
383 3.10.0-693.21.1.el7 (RHEL7.4)
384 3.0.101-0.47.71 (SLES11 SP3)
385 3.0.101-107 (SLES11 SP4)
386 3.12.74-60.64.40 (SLES12 SP1)
387 4.4.114-92.64 (SLES12 SP2)
388 4.4.114-94.11 (SLES12 SP3)
389 3.13.0-101 (Ubuntu 14.04)
390 4.4.0-85.108 (Ubuntu 14.04.5 LTS)
391 4.4.0-116 (Ubuntu 16.04)
393 * Recommended e2fsprogs version: 1.42.13.wc6 or newer
394 * Recommended ZFS / SPL version: 0.7.6
395 * Tested with ZFS / SPL version: 0.7.6
396 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
397 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
398 stack overflow. For more information, please refer to bugzilla 17630.
399 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
400 the re-exporting nodes to be mounted with "32bitapi" mount option
401 * Note for Ubuntu 14.04.5:
402 The 4.4.0 series is the HWE Kernel derived from 16.04 (Xenial).
403 See: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
405 --------------------------------------------------------------------------------
407 2017-07-05 Intel Corporation
409 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
410 for currently supported client and server kernel versions.
411 * Server known to build on patched kernels:
412 2.6.32-431.29.2.el6 (RHEL6.5)
413 2.6.32-504.30.3.el6 (RHEL6.6)
414 2.6.32-573.26.1.el6 (RHEL6.7)
415 2.6.32-642.15.1.el6 (RHEL6.8)
416 2.6.32-696.3.1.el6 (RHEL6.9)
417 3.10.0-514.21.1.el7 (RHEL7.3)
418 3.0.101-0.47.71 (SLES11 SP3)
419 3.0.101-100 (SLES11 SP4)
420 3.12.74-60.64.40 (SLES12 SP1)
421 4.4.59-92.17 (SLES12 SP2)
422 vanilla linux 4.6.7 (ZFS only)
423 * Client known to build on unpatched kernels:
424 2.6.32-431.29.2.el6 (RHEL6.5)
425 2.6.32-504.30.3.el6 (RHEL6.6)
426 2.6.32-573.26.1.el6 (RHEL6.7)
427 2.6.32-642.15.1.el6 (RHEL6.8)
428 2.6.32-696.3.1.el6 (RHEL6.9)
429 3.10.0-514.21.1.el7 (RHEL7.3)
430 3.0.101-0.47.71 (SLES11 SP3)
431 3.0.101-100 (SLES11 SP4)
432 3.12.74-60.64.40 (SLES12 SP1)
433 4.4.59-92.17 (SLES12 SP2)
435 * Recommended e2fsprogs version: 1.42.13.wc4 or newer
436 * Recommended ZFS / SPL version: 0.6.5.9
437 * Tested with ZFS / SPL version: 0.6.5.9
438 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
439 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
440 stack overflow. For more information, please refer to bugzilla 17630.
441 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
442 the re-exporting nodes to be mounted with "32bitapi" mount option
444 --------------------------------------------------------------------------------
446 12-07-2016 Intel Corporation
448 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
449 for currently supported client and server kernel versions.
450 * Server known to build on patched kernels:
451 2.6.32-431.29.2.el6 (RHEL6.5)
452 2.6.32-504.30.3.el6 (RHEL6.6)
453 2.6.32-573.26.1.el6 (RHEL6.7)
454 2.6.32-642.11.1.el6 (RHEL6.8)
455 3.10.0-514.el7 (RHEL7.3)
456 3.0.101-0.47.71 (SLES11 SP3)
457 3.0.101-80 (SLES11 SP4)
458 3.12.62-60.64.8 (SLES12 SP1)
459 vanilla linux 4.5.7 (ZFS only)
460 * Client known to build on unpatched kernels:
461 2.6.32-431.29.2.el6 (RHEL6.5)
462 2.6.32-504.30.3.el6 (RHEL6.6)
463 2.6.32-573.26.1.el6 (RHEL6.7)
464 2.6.32-642.11.1.el6 (RHEL6.8)
465 3.10.0-514.el7 (RHEL7.3)
466 3.0.101-0.47.71 (SLES11 SP3)
467 3.0.101-80 (SLES11 SP4)
468 3.12.62-60.64.8 (SLES12 SP1)
470 * Recommended e2fsprogs version: 1.42.13.wc5 or newer
471 * Recommended ZFS / SPL version: 0.6.5.8
472 * Tested with ZFS / SPL version: 0.6.5.8
473 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
474 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
475 stack overflow. For more information, please refer to bugzilla 17630.
476 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
477 the re-exporting nodes to be mounted with "32bitapi" mount option
479 --------------------------------------------------------------------------------
481 02-29-2016 Intel Corporation
483 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
484 for currently supported client and server kernel versions.
485 * Server known to build on patched kernels:
486 2.6.32-431.29.2.el6 (RHEL6.5)
487 2.6.32-504.30.3.el6 (RHEL6.6)
488 2.6.32-573.12.1.el6 (RHEL6.7)
489 3.10.0-327.3.1.el7 (RHEL7.2)
490 3.0.101-0.47.71 (SLES11 SP3)
491 3.0.101-68 (SLES11 SP4)
492 vanilla linux 4.2.1 (ZFS only)
493 * Client known to build on unpatched kernels:
494 2.6.32-431.29.2.el6 (RHEL6.5)
495 2.6.32-504.30.3.el6 (RHEL6.6)
496 2.6.32-573.12.1.el6 (RHEL6.7)
497 3.10.0-327.3.1.el7 (RHEL7.2)
498 3.0.101-0.47.71 (SLES11 SP3)
499 3.0.101-68 (SLES11 SP4)
502 * Recommended e2fsprogs version: 1.42.13.wc4 or newer
503 * Recommended ZFS / SPL version: 0.6.4.2
504 * Tested with ZFS / SPL version: 0.6.4.2
505 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
506 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
507 stack overflow. For more information, please refer to bugzilla 17630.
508 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
509 the re-exporting nodes to be mounted with "32bitapi" mount option
511 --------------------------------------------------------------------------------
513 03-10-2015 Intel Corporation
515 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
516 for currently supported client and server kernel versions.
517 * Server known to build on patched kernels:
518 2.6.32-431.29.2.el6 (RHEL6.5)
519 2.6.32-504.8.1.el6 (RHEL6.6)
520 3.0.101-0.46 (SLES11 SP3)
521 * Client known to build on unpatched kernels:
522 2.6.32-431.29.2.el6 (RHEL6.5)
523 2.6.32-504.8.1.el6 (RHEL6.6)
524 3.10.0-123.20.1.el7 (RHEL7)
525 3.0.101-0.46 (SLES11 SP3)
526 * Recommended e2fsprogs version: 1.42.9.wc1 or newer
527 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
528 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
529 stack overflow. For more information, please refer to bugzilla 17630.
530 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
531 the re-exporting nodes to be mounted with "32bitapi" mount option
533 Severity : enhancement
535 Description: control OST-index in IDIF via ROCOMPAT flag.
536 Details : Introduce new flag OBD_ROCOMPAT_IDX_IN_IDIF that is stored in the
537 last_rcvd file. For new formatted OST device, it will be auto set;
538 for the case of upgrading from old OST device, you can enable it
539 via the lproc interface osd-ldiskfs.index_in_idif. With such flag
540 enabled, for new created OST-object, its IDIF-in-LMA will contain
541 the OST-index; for the existing OST-object, the OSD will convert
542 old format IDIF as new format IDIF with OST-index stored in the
543 LMA EA when accessing such OST-object or via OI scrub. Once such
544 flag is enabled, it cannot be reverted back, so the system cannot
545 be downgraded to the orignal incompatible version.
547 --------------------------------------------------------------------------------
549 07-30-2014 Intel Corporation
551 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
552 for currently supported client and server kernel versions.
553 * Server known to build on patched kernels:
554 2.6.32-431.20.3.el6 (RHEL6.5)
555 3.0.101-0.31 (SLES11 SP3)
556 * Client known to build on unpatched kernels:
557 2.6.32-431.20.3.el6 (RHEL6.5)
558 3.0.101-0.31 (SLES11 SP3)
561 * Recommended e2fsprogs version: 1.42.9.wc1 or newer
562 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
563 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
564 stack overflow. For more information, please refer to bugzilla 17630.
565 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
566 the re-exporting nodes to be mounted with "32bitapi" mount option
568 --------------------------------------------------------------------------------
570 10-31-2013 Intel Corporation
572 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
573 * Server builds for kernels:
574 2.6.32-358.18.1.el6 (RHEL6.4)
575 * Client builds for unpatched kernels:
576 2.6.32-358.18.1.el6 (RHEL6.4)
577 3.0.92-0.8 (SLES11 SP3)
579 * Recommended e2fsprogs version: 1.42.7.wc2 or newer
580 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
581 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
582 stack overflow. For more information, please refer to bugzilla 17630.
583 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
584 the re-exporting nodes to be mounted with "32bitapi" mount option
586 --------------------------------------------------------------------------------
588 05-30-2013 Intel Corporation
590 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
591 * Server builds for kernels:
592 2.6.32-358.6.2.el6 (RHEL6.4)
593 * Client builds for unpatched kernels:
594 2.6.32-358.6.2.el6 (RHEL6.4)
595 3.0.74-0.6.6 (SLES11 SP2)
597 * Recommended e2fsprogs version: 1.42.7.wc1 or newer
598 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
599 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
600 stack overflow. For more information, please refer to bugzilla 17630.
601 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
602 the re-exporting nodes to be mounted with "32bitapi" mount option
604 --------------------------------------------------------------------------------
606 9-30-2012 Whamcloud, Inc.
608 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
609 * Server support for kernels:
610 2.6.32-279.5.1.el6 (RHEL6)
611 * Client support for unpatched kernels:
612 2.6.32-279.5.1.el6 (RHEL6)
614 2.6.18-238.19.1.el5 (RHEL5)
615 2.6.16.60-0.69.1 (SLES10)
616 * Recommended e2fsprogs version: 1.41.90.wc4
617 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
618 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
619 stack overflow. For more information, please refer to bugzilla 17630.
620 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
621 the re-exporting nodes to be mounted with "32bitapi" mount option
623 Severity : enhancement
625 Description: Update RHEL6.3 kernel to 2.6.32-279.5.1.el6.
627 Severity : enhancement
629 Description: Update RHEL6.3 kernel to 2.6.32-279.2.1.el6.
631 Severity : enhancement
633 Description: Update RHEL6.2 kernel to 2.6.32-220.17.1.el6
635 Severity : enhancement
637 Description: Update RHEL6.2 kernel to 2.6.32-220.13.1.el6
639 Severity : enhancement
641 Description: Update RHEL6.2 kernel to 2.6.32-220.7.1.el6
643 --------------------------------------------------------------------------------
645 04-30-2012 Whamcloud, Inc.
647 * See https://wiki.whamcloud.com/display/PUB/Lustre+Support+Matrix
648 * Server support for kernels:
649 2.6.32-220.4.2.el6 (RHEL6)
650 * Client support for unpatched kernels:
651 2.6.32-220.4.2.el6 (RHEL6)
653 2.6.18-238.19.1.el5 (RHEL5)
654 2.6.16.60-0.69.1 (SLES10)
655 * Recommended e2fsprogs version: 1.41.90.wc4
656 * NFS export disabled when stack size < 8192 (32-bit Lustre clients),
657 since the NFSv4 export of Lustre filesystem with 4K stack may cause a
658 stack overflow. For more information, please refer to bugzilla 17630.
659 * NFSv4 reexport to 32-bit NFS client nodes requires Lustre client on
660 the re-exporting nodes to be mounted with "32bitapi" mount option
662 Severity : enhancement
664 Description: Update RHEL6.2 kernel to 2.6.32-220.4.2.el6
666 Severity : enhancement
668 Description: Update RHEL6.2 kernel to 2.6.32-220.4.1.el6
670 Severity : enhancement
672 Description: Update RHEL6.2 kernel to 2.6.32-220.el6
674 Severity : enhancement
676 Description: Update RHEL6.1 kernel to 2.6.32-131.17.1.el6
678 --------------------------------------------------------------------------------
680 09-30-2011 Whamcloud, Inc.
682 * Server support for kernels:
683 2.6.18-238.19.1.el5 (RHEL5)
684 2.6.32-131.6.1.el6 (RHEL6)
685 * Client support for unpatched kernels:
686 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
687 2.6.18-238.19.1.el5 (RHEL5)
688 2.6.32-131.6.1.el6 (RHEL6)
689 2.6.29.4-167.fc11 (FC11)
690 2.6.16.60-0.69.1 (SLES10)
692 * Recommended e2fsprogs version: 1.41.90.wc3
693 * NFS export disabled when stack size < 8192. Since the NFSv4 export of
694 Lustre filesystem with 4K stack may cause a stack overflow. For more
695 information, please refer to bugzilla 17630.
696 * NFSv4 reexport requires Lustre client on the re-exporting nodes to be
697 mounted with "32bitapi" mount option
699 Severity : enhancement
700 Description: Update RHEL6.1 kernel to 2.6.32-131.6.1.el6
702 Severity : enhancement
703 Description: Update RHEL5.6 kernel to 2.6.18-238.19.1.el5
705 Severity : enhancement
706 Description: Update RHEL5.6 kernel to 2.6.18-238.9.1.el5
709 Frequency : only with SLES10
711 Description: Use OFED "KMP" provided by Novell
712 Details : SLES10 SP3 ships with OFED in a separate "KMP" package.
713 Lustre is now built against this package. That means you need to
714 install the ofed-kmp package from Novell for the patchless client.
715 Note that the ofed-kmp that Novell ships may not exactly match the
716 kernel version but should still be compatible.
718 Severity : enhancement
720 Description: Update RHEL5.5 kernel to 2.6.18-194.17.1.el5
721 Update OEL5.5 kernel to 2.6.18-194.17.1.0.1.el5.
723 --------------------------------------------------------------------------------
725 2010-07-15 Oracle, Inc.
727 * Server support for kernels:
728 2.6.18-164.11.1.el5 (RHEL 5)
729 2.6.18-164.11.1.0.1.el5 (OEL 5)
730 * Client support for unpatched kernels:
731 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
732 2.6.18-164.11.1.el5 (RHEL 5),
733 2.6.18-164.11.1.0.1.el5 (OEL 5)
734 2.6.16.60-0.42.8 (SLES 10),
736 2.6.29.4-167.fc11 (FC11)
737 * Recommended e2fsprogs version: 1.41.10-sun2
739 Severity : enhancement
741 Description: Distribute lfs_migrate script with lustre
742 Details : The lfs_migrate mentioned in the manual is now included in the
743 lustre tree in lustre/scripts and will distributed in lustre RPMs.
745 Severity : enhancement
747 Description: Failover nid list added to import proc file.
749 Severity : enhancement
751 Description: "som_preview" mount option required to use SOM
752 Details : The clients will need to use som_preview mount option
756 Description: Lustre does not do 1MB IOs to HW RAID
757 Details : Bump MAX_PHYS/HW_SEGMENTS and SG_ALL to 256 in the RHEL5 kernel.
759 Severity : enhancement
761 Description: lfs quota output is not very convenient for awk/sed-parsing
762 Details : Some positions in lfs quota output table could be empty or
763 non-empty which made it hard to parse it with scripts, now
764 a dash is put instead of space where there is not supposed
769 Description: Update to OFED 1.5.1
770 Details : OFED 1.5.1 will be used in environments that don't use
771 vendor supplied OFED.
775 Description: Remove lnet. prefix from lctl params display
776 Details : Also renamed lnet.memused to lnet_memused to
781 Description: Use conf_param -d to remove permanant settings.
785 Description: Add list_param -R
786 Details : lctl list_param now supports recursive mode.
789 Bugzilla : 21829, 22312
790 Description: Fix llobdstat.
791 Details : 1. Now updating the write_bytes counter
792 2. Port script fixes from 1.8.
796 Description: decrease the usage of memory on clients.
797 Details : 1. On clients, recycle dentries and inodes unused.
798 2. Delete the code related to ll_deathrow(att 6215 in bug 1443). It
803 Description: llobdstat fix and enhancement
804 Details : add a counter to set a limit to how many samples will be returned
805 fix a wildcard in the path to limit to obdfilter stats only
809 Description: lfs check only allowed for root
810 Details : Code cleanups and added a regresion test of non-root use of
815 Description: MDS Oopses in ptlrpc_dispatch_difficult_reply()/__list_add()
816 Details : Fix hr->hr_index overload.
820 Description: Fix FIEMAP functionality.
824 Description: FC11 patchless client support added
828 Description: Update RHEL5.4 kernel to 2.6.18-164.11.1.el5 and
829 OEL5.4 kernel to 2.6.18-164.11.1.0.1.el5.
833 Description: New acc-sm script and various test-framework enhancments.
837 Description: Previously landed patch was reverted.
841 Description: Update SLES10 kernel to 2.6.16.60-0.42.8 (patchless)
845 Description: Add ldiskfs maxdirsize mount option.
849 Description: Fix mkfs.lustre support for 16TB LUNs.
853 Description: Print mmp_check_interval
854 Details : print mmp_check_interval and make it possible to abort
855 mount operation in case it takes too long.
859 Description: Add cascading_rw to lustre mpi tests.
863 Description: Store SOM data in EA.
864 Details : Also several SOM related cleanups.
868 Description: Add refcount to md_open_data
872 Description: Use refcount on lov_request_set to prevent memory corruption.
876 Description: Statahead: use parent dir's i_mutex for synchronization.
880 Description: loadgen improvements.
884 Description: Enable peer health detection in ksocklnd and ko2iblnd.
888 Description: Control DCACHE_LUSTRE_INVALID flag with MDS_INODE_LOOKCUP lock
892 Description: Fix race on obd_devcie:md_stats
896 Description: Fix SLES 11 client builds.
897 Details : Resolved uninitalized pointer gcc warnings.
901 Description: Allow non-root access for "lfs check".
902 Details : Added a check in obd_class_ioctl() for OBD_IOC_PING_TARGET.
906 Description: Clear imp_force_reconnect correctly in ptlrpc_connect_interpret()
908 Severity : enhancement
910 Description: Update kernel to RHEL5.4 2.6.18-164.6.1.el5 and
911 OEL5 2.6.18-164.6.1.0.1.el5(Both in-kernel OFED enabled).
915 Description: Sanity test_27p fails.
916 Details : OBD_FAIL_OST_ENOSPC has been moved to filter_statfs().
917 Sanity tests 27 were cleaned up and improved.
921 Description: Unify req format on client/servers
922 Details : Use new req_capsule API [almost] everywhere instead of old PTLRPC
923 buffers and swabbers approach..
928 Description: drop join file stuff.
930 Severity : enhancement
932 Description: Add 2.6.27 and SLES11 (patchless client) support
936 Description: Fix in ptlrpc_expire_one_request() to print the signed time
939 Severity : enhancement
941 Description: Build kernels (RHEL5, OEL5 and SLES10/11) using the vendor's own
944 Severity : enhancement
946 Description: Remove set_info(KEY_UNLINKED) from MDS/OSC
948 Severity : enhancement
950 Description: correctly handle big reply message.
951 Details : send LNet event if reply is bigger than buffer and adjust this buffer
956 Description: Drop unnecessary __GFP_NOMEMALLOC flag from filter_get_page()
958 Severity : enhancement
960 Description: Add support for OEL5.
964 Description: Conf-sanity.sh 50g test - deactivated OST should not cause a panic.
966 Severity : enhancement
968 Description: "weak-modules" support
969 Details : Implement "weak-modules" support which enables kernel modules
970 to be used with any kernel that implements the same kABI. In
971 order to achieve this modules are now installed in
972 /lib/modules/$(uname -r)/updates/kernel on all distributions.
976 Description: client could not reconnect to OST because of an active request.
977 Details : The client now retries to reconnect to the same server, if a
978 connect request failed with EBUSY or -EAGAIN. Test case 35b
979 conf-sanity.sh has been created.
983 Description: Don't allow make backward step on assignin osc next id.
984 Details : race between allocation next id and ll_sync thread can be cause
985 of set wrong osc next id and can be kill valid ost objects.
989 Description: Changes in raid5-large-io-rhel5.patch to calculate sectors properly
993 Description: Increase the default BLK_DEF_MAX_SECTORS value for RHEL5 and SLES11
997 Description: Error handling in osc_statfs_interpret() has been improved.
998 Details : Check in osc_statfs_interpret() for EBADR.
1002 Description: Do not send statfs() requests to OSTs disabled by administrator.
1003 Details : Check in lov_prep_statfs_set() for non-NULL ltd_exp.
1005 Severity : enhancement
1007 Description: Added sanity-quota test to for lfs(1) to work with numeric uid/gid.
1009 Severity : enhancement
1011 Description: Update kernel to RHEL5 2.6.18-128.7.1.el5.
1013 Severity : enhancement
1015 Description: Enhancement for lfs(1) command to use numeric uid/gid.
1019 Description: Improved cleanup procedure for conf-sanity test_46a.
1023 Description: Do not start bulk transfer, if client tried to reconnect.
1024 Details : Do not start bulk transfer, if client tried to reconnect,
1025 but failed due to active rpcs. Check for exp_abort_active_req
1026 has been implemented.
1030 Description: Prevent inconsistences between linux and lustre mount structures.
1031 Details : Wait indefinitely in server_wait_finished() until mnt_count drops.
1032 Make the sleep interruptible.
1034 Severity : enhancement
1036 Description: provide server to client comms path
1037 Details : server to client communications path via new LDLM_SET_INFO rpc
1041 Description: truncate starts GFP_FS allocation under transaction causing deadlock
1042 Details : ldiskfs_truncate calls grab_cache_page which may start page
1043 allocation under an open transaction. This may lead to
1044 calling prune_icache with consequent lustre reentrance.
1048 Description: send correctly lsm on open replay
1049 Details : MDS is trust to LSM size on replay open, but client can set wrong size
1054 Description: enable client interface failover
1055 Details : When a child reconnects from another NID, properly update export
1056 nid hash position and ldlm reverse import.
1058 Severity : enhancement
1060 Description: Communicate OST degraded/readonly state via statfs to MDS
1061 Details : Flags in the statfs returned from OSTs indicate whether the
1062 OST is in a degraded RAID state, or if the filesystem has
1063 turned read-only after a filesystem error is detected.
1068 Description: don't panic if EPROTO was hit when reading symlink
1069 Details : correctly handling request reference in error cases.
1071 Severity : enhancement
1073 Description: Add LustreNetLink, a kernel-userspace communcation path.
1075 Severity : enhancement
1077 Description: Update kernel to SLES10 SP2 2.6.16.60-0.39.3.
1082 Description: access to llog context before init.
1083 Details : move handling CATALOGS file at osc layer and forbid access to llog
1084 context before init.
1087 Frequency : rare, only when client is writing to OST on local node
1089 Description: Avoid deadlock for local client writes
1090 Details : Use new OBD_BRW_MEMALLOC flag to notify OST about writes in the
1091 memory freeing context. This allows OST threads to set the
1092 PF_MEMALLOC flag on task structures in order to allocate memory
1093 from reserved pools and complete IO.
1094 Use GFP_HIGHUSER for OST allocations for non-local client writes,
1095 so that the OST threads generate memory pressure and allow
1096 inactive pages to be reclaimed.
1098 Severity : enhancement
1100 Description: Update kernel to RHEL5.3 2.6.18-128.1.14.el5.
1105 Description: lock ordering violation between &cli->cl_sem and _lprocfs_lock
1106 Details : move ldlm namespace creation in setup phase to avoid grab
1107 _lprocfs_lock with cli_sem held.
1111 Description: Temporarily disable grant shrink.
1112 Details : Disable the feature for debugging.
1115 Frequency : only when formatting test filesystems
1117 Description: Unable to run several mkfs.lustre on loop devices at the same time
1118 Details : mkfs.lustre returns error 256 on the concurrent loop devices
1119 formatting. The solution is to proper handle the error.
1121 Severity : enhancement
1123 Description: Update kernel to RHEL5.3 2.6.18-128.1.6.el5.
1125 Severity : enhancement
1127 Description: Update kernel to SLES10 SP2 2.6.16.60-0.37.
1132 Description: resolve race between obd_disconnect and class_disconnect_exports
1133 Details : if obd_disconnect will be called to already disconnected export he
1134 forget release one reference and osc module can't unloaded.
1136 Severity : enhancement
1138 Description: Allow tuning service thread via /proc
1139 Details : For each service a new
1140 /proc/fs/lustre/{service}/*/thread_{min,max,started} entry is
1141 created that can be used to set min/max thread counts, and get the
1142 current number of running threads.
1147 Descriptoin: don't return error if have partially created objects for file.
1148 Details : lov_update_create_set uses set->set_success as index for created
1149 objects, so if some requests will be failed, they will have hole
1150 at end of array and qos_shrink_lsm for allocate correct lsm.
1152 Severity : enhancement
1154 Description: Update OFED support to 1.4.1
1159 Description: Reduce small size read RPC
1160 Details : Set read-ahead limit for every file and only do read-ahead when
1161 available read-ahead pages are bigger than 1M to avoid small size
1164 Severity : enhancement
1166 Description: move AT tunable parameters for more consistent usage
1167 Details : add AT tunables under /proc/sys/lustre, add to conf_param parsing
1169 Severity : enhancement
1171 Description: add lazystatfs mount option to allow statfs(2) to skip down OSTs
1172 Details : allow skip disconnected ost for send statfs request and hide error
1178 Description: fix racy locking of mballoc block bitmaps causing BUG
1179 Details : The locking of the mballoc buddy bitmap and the in-memory
1180 block bitmap was using two different spin locks in some
1181 cases. This made it possible to incorrectly access the
1182 mballoc bitmap while another process was modifying it,
1183 causing a sanity assertion to fail. While no on-disk corruption
1184 was reported, there was some risk of this happening.
1187 Frequency : rare, on llog test 6
1189 Descriptoin: don't allow connect to already connected import
1190 Details : allowing connect to already connected import is hide connecting problem.
1195 Descriptoin: allow kill process which wait statahead result
1196 Details : for some reasons 'ls' can stick in waiting result from statahead,
1197 in this case need way for kill this process.
1199 Severity : enhancement
1201 Description: Add state history info file, enhance import info file
1202 Details : Track import connection state changes in a new osc/mdc proc file;
1203 add overview-type data to the osc/mdc import proc file.
1205 Severity : enhancement
1207 Description: MDS create should not wait for statfs RPC while holding DLM lock.
1209 Severity : enhancement
1211 Description: Update to RHEL5U3 kernel-2.6.18-128.1.1.el5.
1216 Descriptoin: OST grant too much space to client even there are not enough space.
1217 Details : Client will shrink its grant cache to OST if there are no write
1218 activity over 6 mins (GRANT_SHRINK_INTERVAL), and OST will retrieve
1219 this grant cache if there are already not enough avaible space
1220 (left_space < total_clients * 32M).
1223 Frequency : start MDS on uncleanly shutdowned MDS device
1225 Descriptoin: ll_sync thread stay in waiting mds<>ost recovery finished
1226 Details : stay in waiting mds<>ost recovery finished produce random bugs
1227 due race between two ll_sync thread for one lov target. send
1228 ACTIVATE event only if connect realy finished and import have
1232 Frequency : rare, connect and disconnect target at same time
1234 Descriptoin: ASSERTION(atomic_read(&imp->imp_inflight) == 0
1235 Details : don't call obd_disconnect under lov_lock. this long time
1236 operation and can block ptlrpcd which answer to connect request.
1241 Descriptoin: don't lose wakeup for imp_recovery_waitq
1242 Details : recover_import_no_retry or invalidate_import and import_close can
1243 both sleep on imp_recovery_waitq, but we was send only one wakeup
1247 Frequency : always with long access acl
1249 Descriptoin: mds can't pack reply with long acl.
1250 Details : mds don't control size of acl but they limited by reint/getattr
1253 Severity : enhancement
1255 Description: Update to SLES10 kernel-2.6.16.60-0.33.
1257 Severity : enhancement
1259 Description: Update to RHEL5 kernel-2.6.18-92.1.22.el5.
1262 Frequency : start MDS on uncleanly shutdowned MDS device
1264 Descriptoin: aborting recovery hang on MDS
1265 Details : don't throttle destroy RPCs for the MDT.
1268 Frequency : on remount
1270 Description: external journal device not working after the remount
1271 Details : clear dev_rdonly flag for external journal devices in
1277 Description: shutdown vs evict race
1278 Details : client_disconnect_export vs connect request race.
1279 if client will evicted at this time - we start invalidate
1280 thread without referece to import and import can be freed
1284 Frequency : rare, need acl's on inode.
1286 Description: client can't handle ost additional correctly
1287 Details : if ost was added after client connected to mds client can have
1288 hit lnet_try_match_md ... to big messages to wide striped files.
1289 in this case need teach client to handle config events about add
1290 lov target and update client max ea size at that event.
1292 Severity : enhancement
1294 Description: Changelogs
1295 Details : Changelogs are a lightweight mechanism to track filesystem
1296 metadata and namespace changes. The changelog is recorded
1297 permanently on the MDTs, and is periodically "consumed" / purged
1298 when records are no longer needed.
1300 Severity : enhancement
1302 Description: compact fld format with extents
1303 Details : Store range of seq rather than every seq in FLD. Seq
1304 controller update FLD rather than clients. In Case of CMD, mdt0
1305 has FLD, all other metadata server act as non persistent proxy
1306 for FLD queries and cache fld entries in fld cache.
1311 Description: don't skip ost target if they assigned to file
1312 Details : Drop slow OSCs if we can, but not for requested start idx.
1313 This means "if OSC is slow and it is not the requested
1314 start OST, then it can be skipped, otherwise skip it only
1315 if it is inactive/recovering/out-of-space.
1319 Description: more cleanup in mds_lov
1320 Details : not send LOV EA under replay, we can't know about they size at this
1321 time. Don't allow client connect to mds before any ost connected,
1322 for avoid problems with LOV EA size and returning EIO to client.
1324 Severity : enhancement
1326 Description: Interoperability at server side (Disk interoperability)
1328 Severity : enhancement
1330 Description: Update to RHEL5 kernel-2.6.18-92.1.17.el5.
1332 Severity : enhancement
1334 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.31.
1336 Severity : enhancement
1338 Description: New client IO stack (CLIO).
1340 Severity : enhancement
1342 Description: Commit on sharing. Eliminate inter-client dependencies between
1343 uncommitted transactions by doing transaction commits.
1344 Thereby clients may recovery independently.
1347 Frequency : Create a symlink file with a very long name
1349 Description: ldlm_cancel_pack()) ASSERTION(max >= dlm->lock_count + count)
1350 Details : If there is no extra space in the request for early cancels,
1351 ldlm_req_handles_avail() returns 0 instead of a negative value.
1353 Severity : enhancement
1355 Description: Add /proc entry for import status
1356 Details : The mdc, osc, and mgc import directories now have
1357 an import directory that contains useful import data for debugging
1358 connection problems.
1360 Severity : enhancement
1362 Description: Re-disable certain /proc logging
1363 Details : Enable and disable client's offset_stats, extents_stats and
1364 extents_stats_per_process stats logging on the fly.
1367 Frequency : Only on FC kernels 2.6.22+
1369 Description: oops in statahead
1370 Details : Do not drop reference count for the dentry from VFS when lookup,
1371 VFS will do that by itself.
1373 Severity : enhancement
1375 Description: Generic /proc file permissions
1376 Details : Set /Proc file permissions in a more generic way to enable non-
1377 root users operate on some /proc files.
1381 Description: Hitting mdc_commit_close() ASSERTION
1382 Details : Properly handle request reference release in
1383 ll_release_openhandle().
1387 Description: quota recovery deadlock during mds failover
1388 Details : This patch includes att18982, att18236, att18237 in bz14840.
1390 1. fix osts hang when mds does failover with quotaon
1391 2. prevent watchdog storm when osts threads wait for the
1396 Frequency : only patchless client
1397 Description: add workaround for race between add/remove dentry from hash
1399 Severity : enhancement
1401 Description: Allow OST glimpses to return PW locks
1405 Description: LBUG when llog conf file is full
1406 Details : When llog bitmap is full, ENOSPC should be returned for plain log.
1410 Description: Prevent import from entering FULL state when server in recovery
1414 Description: service mount cannot take device name with ":"
1415 Details : Only when device name contains ":/" will mount treat it as
1421 Description: replace ptlrpcd with the statahead thread to interpret the async
1422 statahead RPC callback
1426 Frequency : on recovery
1427 Description: I/O failures after umount during fail back
1428 Details : if client reconnected to restarted server we need join to recovery
1429 instead of find server handler is changed and process self
1430 eviction with cancel all locks.
1432 Severity : enhancement
1434 Description: Update to RHEL5 kernel-2.6.18-92.1.10.el5.
1436 Severity : enhancement
1438 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.27.
1440 Severity : enhancement
1442 Description: Upcall on Lustre log has been dumped
1443 Details : Allow for a user mode script to be called once a Lustre log has
1444 been dumped. It passes the filename of the dumped log to the
1445 script, the location of the script can be specified via
1446 /proc/sys/lnet/debug_log_upcall.
1451 Description: avoid idr_remove called for id which is not allocated.
1452 Details : Move assigment s_dev for clustered nfs to end of initialization,
1453 to avoid problem with error handling.
1458 Description: avoid Already found the key in hash [CONN_UNUSED_HASH] messages
1459 Details : When connection is reused this not moved from CONN_UNUSED_HASH
1460 into CONN_USED_HASH and this prodice warning when put connection
1461 again in unused hash.
1464 Severity : enhancement
1466 Description: File striping can now be set to use an arbitrary pool of OSTs.
1468 Severity : enhancement
1470 Description: Export bytes_read/bytes_write count on OSC/OST.
1474 Description: Early reply size mismatch, MGC loses connection
1475 Details : Apply the MGS_CONNECT_SUPPORTED mask at reconnect time so
1476 the connect flags are properly negotiated.
1481 Description: quotas are not honored with O_DIRECT
1482 Details : all writes with the flag O_DIRECT will use grants which leads to
1483 this problem. Now using OBD_BRW_SYNC to guard this.
1487 Description: add quota statistics
1488 Details : 1. sort out quota proc entries and proc code.
1489 2. add quota statistics
1491 Severity : enhancement
1493 Description: enable quota support for HEAD.
1497 Description: Properly propagate oinfo flags from lov to osc for statfs
1498 Details : restore missing copy oi_flags to lov requests.
1500 Severity : enhancement
1502 Description: Add man pages for llobdstat(8), llstat(8), plot-llstat(8),
1503 : l_getgroups(8), lst(8), routerstat(8)
1504 Details : included man pages for llobdstat(8), llstat(8),
1505 : plot-llstat(8), l_getgroups(8), lst(8), routerstat(8)
1507 Severity : enhancement
1509 Description: configure's --enable-quota should check the
1510 : kernel .config for CONFIG_QUOTA
1511 Details : configure is terminated if --enable-quota is passed but
1512 : no quota support is in kernel
1516 Description: Remove portals compatibility
1517 Details : Remove portals compatibility, not interoperable with releases
1522 Description: Resolve device initialization race
1523 Details : Prevent proc handler from accessing devices added to the
1524 obd_devs array but yet be intialized.
1526 Severity : enhancement
1528 Description: Update to SLES10 SP2 kernel-2.6.16.60-0.23.
1530 Severity : enhancement
1532 Description: Update to RHEL5 kernel-2.6.18-92.1.6.el5.
1537 Description: Using wrong pointer in osc_brw_prep_request
1538 Details : Access to array[-1] can produce panic if kernel compiled with
1539 CONFIG_PAGE_ALLOC enabled
1543 Description: Client runs out of low memory
1544 Details : Consider only lowmem when counting initial number of llap pages
1548 Description: *optional* service tags registration
1549 Details : if the "service tags" package is installed on a Lustre node
1550 When the filesystem is mounted, a local-node service tag will
1551 be created. See http://inventory.sun.com/ for more information
1552 about the Service Tags asset management system.
1556 Description: Kernel BUG tries to release flock
1557 Details : Lustre does not destroy flock lock before last reference goes
1558 away. So always drop flock locks when client is evicted and
1559 perform unlock regardless of successfulness of speaking to MDS.
1563 Description: add refcount protection for osc callbacks, avoid panic on shutdown
1567 Description: sanity test 65a fails if stripecount of -1 is set
1568 Details : handle -1 striping on filesystem in ll_dirstripe_verify
1573 Description: ASSERTION(CheckWriteback(page,cmd)) failed
1574 Details : badly clear PG_Writeback bit in ll_ap_completion can produce false
1577 Severity : enhancement
1579 Description: Update to RHEL5 kernel-2.6.18-53.1.21.el5.
1583 Description: do not process already freed flock
1584 Details : flock can possibly be freed by another thread before it reaches
1585 to ldlm_flock_completion_ast.
1589 Description: LBUG during stress test
1590 Details : Need properly lock accesses the flock deadlock detection list.
1594 Description: oops in page fault handler
1595 Details : kernel page fault handler can return two special 'pages' in error
1596 case, don't try dereference NOPAGE_SIGBUS and NOPAGE_OMM.
1600 Description: timeout with invalidate import.
1601 Details : ptlrpcd_check call obd_zombie_impexp_cull and wait request which
1602 should be handled by ptlrpcd. This produce long age waiting and
1603 -ETIMEOUT ptlrpc_invalidate_import and as result LASSERT.
1605 Severity : enhancement
1607 Description: Update to RHEL5 kernel-2.6.18-53.1.19.el5.
1611 Description: enable MGS and MDT services start separately
1612 Details : add a 'nomgs' option in mount.lustre to enable start a MDT with
1613 a co-located MGS without starting the MGS, which is a complement
1614 to 'nosvc' mount option.
1618 Frequency : after recovery
1619 Description: precreate to many object's after del orphan.
1620 Details : del orphan st in oscc last_id == next_id and this triger growing
1621 count of precreated objects. Set flag LOW to skip increase count
1622 of precreated objects.
1626 Frequency : rare, on clear nid stats
1627 Description: ASSERTION(client_stat->nid_exp_ref_count == 0)
1628 Details : when clean nid stats sometimes try destroy live entry,
1629 and this produce panic in free.
1633 Description: Stack overflow during MDS log replay
1634 ease stack pressure by using a thread dealing llog_process.
1638 Description: wait until IO finished before start new when do lock cancel.
1639 Details : VM protocol want old IO finished before start new, in this case
1640 need wait until PG_writeback is cleared until check dirty flag and
1641 call writepages in lock cancel callback.
1643 Severity : enhancement
1645 Description: using special macro for print time and cleanup in includes.
1649 Description: mds_mfd_close() ASSERTION(rc == 0)
1650 Details : In mds_mfd_close(), we need protect inode's writecount change
1651 within its orphan write semaphore to prevent possible races.
1655 Description: Obsolete CURRENT_SECONDS and use cfs_time_current_sec() instead.
1659 Frequency : rare, on shutdown ost
1660 Description: don't hit live lock with umount ost.
1661 Details : shrink_dcache_parent can be in long loop with destroy dentries,
1662 use shrink_dcache_sb instead.
1666 Description: don't panic with use echo client
1667 Details : echo client pass NULL as client nid pointer and this produce null
1668 pointer dereference.
1672 Description: fix build on ppc32
1673 Details : compile code with -m64 flag produce wrong object file for ppc32.
1677 Description: add message levels for liblustreapi
1681 Description: fix for occasional failure case of -ENOSPC in recovery-small tests
1682 Details : Move the 'good_osts' check before the 'total_bavail' check. This
1683 will result in an -EAGAIN and in the exit call path we call
1684 alloc_rr() which will with increasing aggressiveness attempt to
1685 aquire precreated objects on the minimum number of required OSCs.
1689 Description: Use old size assignment to avoid deadlock
1690 Details : Reverts the changes in bugs 2369 and bug 14138 that introduced
1691 the scheduling while holding a spinlock. We do not need locking
1692 for size in ll_update_inode() because size is only updated from
1693 the MDS for directories or files without objects, so there is no
1694 other place to do the update, and concurrent access to such inodes
1695 are protected by the inode lock.
1699 Description: resolve "_IOWR redefined" build error on SLES10
1703 Description: dump the memory debugging after all modules are unloaded to
1704 suppress false negative in conf_sanity test 39
1706 Severity : enhancement
1708 Description: build kernel-ib packages for OFED 1.3 in our release cycle
1713 Description: fix SLES kernel versioning
1714 Details : the kernel version for our SLES 10 kernel did not include a "-"
1715 before the "smp" at the end. while this was not a problem in
1716 general, it did mean that software trying to use the kernel
1717 version to try to detect a vendor specific kernel would fail.
1718 this was most evident by the OFED build scripts.
1722 Description: Don't update lov_desc members until making sure they are valid
1723 Details : When updating lov_desc members via proc fs, need fix their
1724 validities before doing the real update.
1728 Description: don't put request into delay list while invalidate in flight.
1729 Details : ptlrpc_delay_request sometimes put in delay list while invalidate
1730 import in flight. this produce timeout for invalidate and sometimes
1731 can cause stale data.
1735 Frequency : on ppc only
1736 Description: not convert ost objects for directory because it's not exist.
1737 Details : ll_dir_getstripe assume dirrectory has ost objects but this wrong.
1741 Description: Add FMODE_EXEC file flag for SLES10 SP1 kernel.
1743 Severity : enhancement
1745 Description: Update to support 2.6.22.14 vanilla kernel.
1749 Frequency : rare, on recovery
1750 Description: read procfs can produce deadlock in some situation
1751 Details : Holding lprocfs lock which send rpc can produce block for destroy
1752 obd objects and this also block reconnect with -EALREADY. This
1753 isn't fix all lprocfs bugs - but make it rare.
1755 Severity : enhancement
1757 Description: Update kernel to RHEL5 2.6.18-53.1.14.el5.
1760 Frequency : frequent on X2 node
1762 Description: mdc_set_open_replay_data LBUG
1763 Details : Set replay data for requests that are eligible for replay.
1767 Description: lustre_mgs: operation 101 on unconnected MGS
1768 Details : When MGC is disconnected from MGS long enough, MGS will evict the
1769 MGC, and late on MGC cannot successfully connect to MGS and a lot
1770 of the error messages complaining that MGS is not connected.
1774 Frequency : on network error
1775 Description: panic with double free request if network error
1776 Details : mdc_finish_enqueue is finish request if any network error ocuring,
1777 but it's true only for synchronus enqueue, for async enqueue
1778 (via ptlrpcd) this incorrect and ptlrpcd want finish request
1781 Severity : enhancement
1783 Description: client-side metadata stat-ahead during readdir(directory readahead)
1784 Details : perform client-side metadata stat-ahead when the client detects
1785 readdir and sequential stat of dir entries therein
1788 Frequency : on start mds
1790 Description: Implement get_info(last_id) in obdfilter.
1793 Frequency : occasional
1795 Description: Correctly check stale fid, not start epoch if ost not support SOM
1796 Details : open with flag O_CREATE need set old fid in op_fid3 because
1797 op_fid2 was overwritten with new generated fid, but mds can answer
1798 with one of these two fids and both is not stale. Setattr
1799 incorrectly started an epoch and assume will be called
1800 done_writing, but without SOM done_writing ever being called.
1803 Frequency : rare, depends on device drivers and load
1805 Description: MDS or OSS nodes crash due to stack overflow
1806 Details : Code changes in 1.8.0 increased the stack usage of some functions.
1807 In some cases, in conjunction with device drivers that use a lot
1808 of stack the MDS (or possibly OSS) service threads could overflow
1809 the stack. One change which was identified to consume additional
1810 stack has been reworked to avoid the extra stack usage.
1813 Frequency : occasional
1815 Description: Do not fail import if osc_interpret_create gets -EAGAIN
1816 Details : If osc_interpret_create got -EAGAIN it immediately exits and
1817 wakeup oscc_waitq. After wakeup oscc_wait_for_objects call
1818 oscc_has_objects and see OSC has no objests and call
1819 oscc_internal_create to resend create request.
1821 Severity : enhancement
1823 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.5.
1825 Severity : enhancement
1827 Description: Update to RHEL5 latest kernel-2.6.18-53.1.13.el5.
1830 Frequency : very rare
1832 Description: Fix replay if there is an un-replied request and open
1833 Details : In some cases, older replay request will revert the
1834 mcd->mcd_last_xid on MDS which is used to record the client's
1835 latest sent request.
1837 Severity : enhancement
1839 Description: Update to RHEL5 latest kernel-2.6.18-53.1.6.el5.
1841 Severity : enhancement
1843 Description: Add rhel5 support to HEAD.
1845 Serverity : enhancement
1847 Description: Update RHEL4 kernel to 2.6.9-67.0.4.
1852 Description: Don't allow skipping OSTs if index has been specified.
1853 Details : Don't allow skipping OSTs if index has been specified, make
1854 locking in internal create lots better.
1858 Description: LBUG in ptlrpc_check_set() bad phase ebc0de00
1859 Details : access to bitfield in structure is always rounded to long
1860 and this produce problem with not atomic change any bit.
1864 Description: Lustre make rpms failed.
1865 Details : Remove ldiskfs spec file to avoids rpmbuild be confused when
1866 builds Lustre rpms from tarball.
1869 Frequency : rare on shutdown ost
1871 Description: If llog cancel was not send before clean_exports phase, this can
1872 produce deadlock in llog code.
1873 Details : If llog thread has last reference to obd and call class_import_put
1874 this produce deadlock because llog_cleanup_commit_master wait when
1875 last llog_commit_thread exited, but this never success because was
1876 called from llog_commit_thread.
1880 Description: allow userland application know is lost one of stripes.
1881 Details : fill lvb_blocks with error code on ost and return it to
1882 application if error flag found.
1886 Description: NULL lov_tgts causing MDS oops
1887 Details : more safe checks for NULL lov_tgts for avoid oops.
1889 Severity : enhancement
1891 Description: Update to RHEL4 latest kernel-2.6.9-67.0.1.EL.
1895 Descriptoin: make lov_create() will not stuck in obd_statfs_rqset()
1896 Details : If an OST is down the MDS will hang indefinitely in
1897 obd_statfs_rqset() waiting for the statfs data. While for
1898 MDS QOS usage of statfs, it should not stuck in waiting.
1900 Severity : enhancement
1902 Description: remote_acl support
1903 Details : Support ACL-based permission check for remote user.
1904 Support setfacl/getfacl for remote user with the utils
1905 "lfs {l,r}{s,g}etfacl" which follow the same parameter format as
1906 the system "{s,g}etfacl" utils.
1908 Severity : enhancement
1910 Description: Update to RHEL4 U6 kernel-2.6.9-67.EL.
1912 Severity : enhancement
1914 Description: Update to RHEL5 latest kernel-2.6.18-53.1.4.el5.
1918 Description: make mgs_setparam() handle fsname containing dash
1919 Details : fsname containing a dash does not work with lctl conf_param
1921 Severity : enhancement
1923 Description: Update to SLES10 SP1 latest kernel-2.6.16.54-0.2.3.
1925 Severity : enhancement
1927 Description: Update to RHEL5 Update-1 kernel 2.6.18-53.el5.
1931 Frequency : rare, at shutdown
1932 Description: access already free / zero obd_namespace.
1933 Details : if client_disconnect_export was called without force flag set,
1934 and exist connect request in flight, this can produce access to
1935 NULL pointer (or already free pointer) when connect_interpret
1936 store ocd flags in obd_namespace.
1940 Frequency : only at startup
1941 Description: not alloc memory with spinlock held.
1942 Details : allocation memory with GFP_KERNEL can produce sleep deadlock,
1943 if any spinlock held.
1945 Severity : enhancement
1947 Description: make lustre randomly fail allocating memory
1948 Details : Make lustre randomly failed allocating memory for testing purpose.
1950 Severity : enhancement
1952 Description: lost problems with lov objid file
1953 Details : Fixes some scability and access to not inited memory problems
1954 in work with lov objdid file.
1959 Description: lfs find does not continue on file error
1960 Details : Continue other files processing when a file/dir is absent.
1964 Description: Inconsistent usage of lustre_pack_reply()
1965 Details : Standardize the usage of lustre_pack_reply() such that it
1966 always generate a CERROR on failure.
1969 Frequency : occasional
1971 Description: MDS hang or stay in waiting lock
1972 Details : If client receive lock with CBPENDING flag ldlm need send lock
1973 cancel as separate rpc, to avoid situation when cancel request
1974 can't processed due all i/o threads stay in wait lock.
1978 Description: Update to RHEL5 kernel 2.6.18-8.1.15.el5.
1982 Description: Update to SLES10 SP1 kernel 2.6.16.53-0.16
1986 Description: Update to SLES9 kernel-2.6.5-7.287.3.
1990 Description: Update to RHEL4 latest kernel.
1992 Severity : enhancement
1994 Description: Build SLES10 patchless client fails
1995 Details : The configure was broken by run ./configure with
1996 --with-linux-obj=.... argument for patchless client. When the
1997 configure use --with-linux-obj, the LINUXINCLUDE= -Iinclude
1998 can't search header adequately. Use absolute path such as
1999 -I($LINUX)/include instead.
2003 Description: interrupt oig_wait produce painc on resend.
2004 Details : brw_redo_request can be used for resend requests from ptlrpcd and
2005 private set, and this produce situation when rq_ptlrpcd_data not
2006 copyed to new allocated request and triggered LBUG on assert
2007 req->rq_ptlrpcd_data != NULL. But this member used only for wakeup
2008 ptlrpcd set if request is changed and can be safety changed to use
2013 Description: LASSERT_{REQ,REP}SWAB macros are buggy
2014 Details : If SWAB_PARANOIA is disabled, the LASSERT_REQSWAB and
2015 LASSERT_REPSWAB macros become no-ops, which is incorrect. Drop
2016 these macros and replace them with their difinitions instead.
2020 Description: Update kernel patches for SLES10 2.6.16.53-0.8.
2021 Details : Update which_patch & target file for SLES10 latest kernel.
2024 Description: Remove client patches from SLES 10 kernel.
2025 Details : This causes SLES 10 clients to behave as patchless clients
2026 even on a Lustre-patched (server) kernel.
2028 Severity : enhancement
2030 Description: self-adjustable client's lru lists
2031 Details : use adaptive algorithm for managing client cached locks lru
2032 lists according to current server load, other client's work
2033 pattern, memory activities, etc. Both, server and client
2034 side namespaces provide number of proc tunables for controlling
2037 Severity : enhancement
2039 Description: light-weight GSS support
2040 Details : Support krb5n and krb5a mode, which keep Kerberos 5 authentication
2041 and reduce performance overhead.
2043 Severity : enhancement
2045 Description: Linux keyring support
2046 Details : Support using service of Linux keyring for Lustre GSS internal
2047 context refresh/cache mechanism.
2051 Description: Fix errors in lfs documentation
2052 Details : Fixes man pages
2056 Description: don't use GFP_* in generic Lustre code.
2057 Details : Use cfs_alloc_* functions and CFS_* flags for code portability.
2061 Description: obdclass is limited by single OBD_ALLOC(idarray)
2062 Details : replace OBD_ALLOC/OBD_FREE with OBD_VMALLOC/OBD_VFREE
2066 Description: warnings with build patchless client with vanila 2.6.19 and up
2067 Details : change old ctl_table style and replace ctl_table/ctl_table_header
2068 with cfs_sysctl_table_t/cfs_sysctl_table_header_t
2072 Frequency : Only for SLES
2073 Description: sanity_quota fail test_1
2074 Details : There are multiple occurences of $TSTUSR in SLES's /etc/group
2075 file, which makes TSTID[2] inunique.
2079 Frequency : Only for SLES9
2080 Description: Kernel patches for SLES9 2.6.5-7.286 kernel
2081 Details : Update target/ChangeLog/which_patch .
2085 Frequency : Only for test_10 in sanity-quota.sh
2086 Description: a bug in quota test code
2087 Details : checking a test flag in a wrong place
2091 Frequency : Only for test_13 in sanity-quota.sh
2092 Description: a bug in quota test script
2093 Details : shouldn't check overfull allocated quota
2097 Description: jbd statistics
2098 Details : Port older jbd statistics patch for sles10
2100 Severity : enhancement
2102 Description: Add printing inode info into message about error in writepage.
2107 Description: reply_lock_interpret crash due to race with it and lock cancel.
2108 Details : Do not replay locks that are being cancelled. Do not reference
2109 locks by their address during replay, just by their handle.
2113 Frequency : When flocks are used.
2114 Description: assertion failure in ldlm_cli_enquque_fini for non NULL lock.
2115 Details : Flock locks might destroy just granted lock if it could be merged
2116 with another existing flock, this is done in completion handler,
2117 so teach ldlm_cli_enquque_fini that this is a valid case for
2123 Description: Oops in read and write path when failing to allocate lock.
2124 Details : Check if lock allocation failed and return error back.
2128 Description: lstripe command fails for valid OST index
2129 Details : The stripe offset is compared to lov->desc.ld_tgt_count
2130 instead of lov->desc.ld_active_tgt_count.
2134 Description: sanity.sh failed test 103
2135 Details : RHEL mis-interpret setfacl "-X" param, so we won't test setfacl
2140 Description: df doesn't work properly if diskfs blocksize != 4K
2141 Details : Choose biggest blocksize of OST's as the LOV's blocksize.
2145 Description: lfs find on -1 stripe looping in lsm_lmm_verify_common()
2146 Details : Avoid lov_verify_lmm_common() on directory with -1 stripe count.
2148 Severity : enhancement
2150 Description: Adaptive timeouts
2151 Details : RPC timeouts adapt to changing server load and network
2152 conditions to reduce resend attempts and improve recovery time.
2156 Description: llapi_file_create() does not allow some changes
2157 Details : add llapi_file_open() that allows specifying the mode and
2158 open flags, and also returns an open file handle.
2162 Description: merge and cleanup kernel patches.
2163 Details : Remove mnt_lustre_list in vfs_intent-2.6-rhel4.patch.
2167 Description: Add journal checksum support.(Kernel part)
2168 Details : The journal checksum feature adds two new flags i.e
2169 JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT and
2170 JBD2_FEATURE_COMPAT_CHECKSUM. JBD2_FEATURE_CHECKSUM flag
2171 indicates that the commit block contains the checksum for
2172 the blocks described by the descriptor blocks. Now commit
2173 record can be sent to disk without waiting for descriptor
2174 blocks to be written to disk. This behavior is controlled
2175 using JBD2_FEATURE_ASYNC_COMMIT flag.
2179 Description: OSS needs mutliple precreate threads
2180 Details : Add ability to start more than one create thread per OSS.
2184 Description: Sanity.sh test_65k failed: setstripe should have succeeded
2185 Details : execute lfs setstripe on client
2189 Description: mds_obd_create error creating tmp object
2190 Details : When the user sets quota on root, llog will be affected and can't
2191 create files and write files.
2194 Frequency : Always on ia64 patchless client, and possibly others.
2196 Description: Add EXPORT_SYMBOL check for node_to_cpumask symbol.
2197 Details : This allows the patchless client to be loaded on architectures
2198 without this export.
2202 Description: RedHat Update kernel for RHEL5
2203 Details : Add the kernel config file for RHEL5.
2207 Description: RedHat Update kernel for RHEL5
2208 Details : Modify the kernel config file more closer RHEL5.
2212 Description: Build failure against Centos5 (RHEL5)
2213 Details : Define PAGE_SIZE when it isn't present.
2217 Description: test_7 on sanity-quota.sh can't be used on separate machines
2218 Details : add facet to handle it
2222 Description: "ll_intent_file_open()) lock enqueue: err: -13" with nfs
2223 Details : with NFS, the anon dentry's parent was set to itself in
2224 d_alloc_anon(), so in MDS, we use rec->ur_fid1 to find the
2225 corresponding dentry other than use rec->ur_name.
2227 Severity : enhancement
2229 Description: lfs setstripe enhancement
2230 Details : Make lfs setstripe understand 'k', 'm' and 'g' for stripe size.
2234 Description: enable data checksumming by default
2235 Details : enable checksum by default, allow --disable-checksum
2236 configure option and "-o nochecksum" mount option. Checksums
2237 can also be disabled at runtime via $LPROC/osc/*/checksum_pages.
2241 Description: lustre support for RHEL5
2242 Details : Add support for RHEL5.
2246 Description: Client eviction due to failover config
2247 Details : after a connection loss, the lustre client should attempt to
2248 reconnect to the last active server first before trying the
2249 other potential connections.
2253 Description: when mds and osts use different quota unit(32bit and 64bit),
2254 quota will be released repeatly.
2255 Details : void sending multiple quota reqs to mds, which will keep the
2256 status between the reqs.
2260 Description: osts not allocated evenly to files
2261 Details : change the condition to increase offset_idx
2264 Frequency : Always for filesystems larger than 2TB on 32-bit systems.
2266 Description: Data corruption for OSTs that are formatted larger than 2TB
2268 Details : When generating the bio request for lustre file writes the
2269 sector number would overflow a temporary variable before being
2270 used for the IO. The data reads correctly from Lustre (which
2271 will overflow in a similar manner) but other file data or
2272 filesystem metadata may be corrupted in some cases.
2276 Description: Tune the kernel for good SCSI performance.
2277 Details : Set the value of /sys/block/{dev}/queue/max_sectors_kb
2278 to the value of /sys/block/{dev}/queue/max_hw_sectors_kb
2283 Description: rewrite ext2-derived code in obdclass/uuid.c
2284 Details : rewrite inherited code (uuid parsing code from ext2 utils)
2285 from scratch preserving functionality.
2289 Description: "lfs find -obd UUID" prints directories
2290 Details : "lfs find -obd UUID" will return all directory names instead
2291 of just file names. It is incorrect because the directories
2292 do not reside on the OSTs.
2296 Description: use i_size_read and i_size_write in 2.6 port
2297 Details : replace inode->i_size access with i_size_read/write()
2300 Frequency : when removing large files
2302 Description: scheduling issue during removal of large Lustre files
2303 Details : Don't take the BKL in fsfilt_ext3_setattr() for 2.6 kernels.
2304 It causes scheduling issues when removing large files (17TB in the
2308 Frequency : only with liblustre clients on XT3
2310 Description: evictions taking too long
2311 Details : allow llrd to evict clients directly on OSTs
2314 Frequency : only on ppc
2316 Description: /proc/fs/lustre/devices broken on ppc
2317 Details : The patch as applied to 1.6.2 doesn't look correct for all arches.
2318 We should make sure the type of 'index' is loff_t and then cast
2319 explicitly as needed below. Do not assign an explicitly cast
2323 Frequency : only for rhel5
2325 Description: Kernel patches update for RHEL5 2.6.18-8.1.10.el5.
2326 Details : Modify the target file & which_kernel.
2328 Severity : enhancement
2330 Description: omit set fsid for export NFS
2331 Details : fix set/restore device id for avoid EMFILE error and mark lustre
2332 fs as FS_REQUIRES_DEV for avoid problems with generate fsid.
2336 Frequency : Always, for kernels after 2.6.16
2337 Description: Fix warning idr_remove called for id=.. which is not allocated.
2338 Details : Last kernels save old s_dev before kill super and not allow
2339 to restore from callback - restore it before call kill_anon_super.
2343 Description: buffer overruns could theoretically occur
2344 Details : llapi_semantic_traverse() modifies the "path" argument by
2345 appending values to the end of the origin string, and a
2346 overrun may occur. Adding buffer overrun check in liblustreapi.
2350 Description: Fix error on 'ls .' at the top of the Lustre mount.
2351 Details : Don't revalidate dentry if it is a root dentry.
2355 Description: Kernel patches update for RHEL4 2.6.9-55.0.6.
2356 Details : Modify vm-tunables-rhel4.patch.
2360 Description: Kernel config for 2.6.18-vanilla.
2361 Details : Modify targets/2.6-vanilla.target.in.
2362 Add config file kernel-2.6.18-2.6-vanilla-i686.config.
2363 Add config file kernel-2.6.18-2.6-vanilla-i686-smp.config.
2364 Add config file kernel-2.6.18-2.6-vanilla-x86_64.config.
2365 Add config file kernel-2.6.18-2.6-vanilla-x86_64-smp.config.
2369 Description: improve handling recoverable errors
2370 Details : if request processig with error which can be recoverable on server
2371 request should be resend, otherwise page released from cache and
2376 Description: Kernel patches update for RHEL5 2.6.18-8.1.14.el5.
2377 Details : Modify target file & which_patch.
2378 A flaw was found in the IA32 system call emulation provided
2379 on AMD64 and Intel 64 platforms. An improperly validated 64-bit
2380 value could be stored in the %RAX register, which could trigger an
2381 out-of-bounds system call table access. An untrusted local user
2382 could exploit this flaw to run code in the kernel
2383 (ie a root privilege escalation). (CVE-2007-4573).
2387 Description: O_DIRECT bypasses client statistics.
2388 Details : When running with O_DIRECT I/O, neither the client rpc_stats nor
2389 read_ahead_stats were updated. copied the stats section from
2390 osc_send_oap_rpc() into async_internal().
2394 Description: Add jbd statistics patch for RHEL5 and 2.6.18-vanilla
2398 Description: change order of libsysio includes
2399 Details : '#include sysio.h' should always come before '#include xtio.h'
2403 Description: handle "serious error: objid * already exists" more gracefully
2404 Details : If LAST_ID value on disk is smaller than the objects existing in
2405 the O/0/d* directories, it indicates disk corruption and causes an
2406 LBUG(). If the object is 0-length, then we should use the existing
2407 object. This will help to avoid a full fsck in most cases.
2409 Severity : enhancement
2411 Description: adapt the lustre_config script to support the upgrade case
2412 Details : Add "-u" option for lustre_config script to support upgrading 1.4
2413 server targets to 1.6 in parallel.
2417 Description: To avoid grant space > avaible space when the disk is almost
2418 full. Without this patch you might see the error "grant XXXX >
2419 available" or some LBUG about grant, when the disk is almost
2421 Details : In filter_check_grant, for non_grant cache write, we should
2422 check the left space by if (*left > ungranted + bytes), instead
2423 of (*left > ungranted), because only we are sure the left space
2424 is enough for another "bytes", then the ungrant space should be
2425 increase. In client, we should update cl_avail_grant only there
2426 is OBD_MD_FLGRANT in the reply.
2430 Description: Update RHEL 4 kernel to fix local root privilege escalation.
2431 Details : Update to the latest RHEL 4 kernel to fix the vulnerability
2432 described in CVE-2007-4573. This problem could allow untrusted
2433 local users to gain root access.
2436 Frequency : when using O_DIRECT and quotas
2438 Description: Incorrect file ownership on O_DIRECT output files
2439 Details : block usage reported by 'lfs quota' does not take into account
2440 files that have been written with O_DIRECT.
2445 Description: touch file failed when fs is not full
2446 Details : OST in recovery should not be discarded by MDS in alloc_qos(),
2447 otherwise we can get ENOSP while fs is not full.
2451 Description: parallel lock callbacks
2452 Details : Instead of sending blocking and completion callbacks as separated
2453 requests, adding them to a set and sending in parallel.
2456 Frequency : only for Cray XT3
2457 Bugzilla : 12829/13455
2458 Description: Changing primary group doesn't change the group lustre assigns to
2460 Details : When CRAY_XT3 is defined, the fsgid supplied by the client is
2461 overridden with the primary group provided by the group upcall,
2462 whereas the supplied fsgid can be trusted if it is in the list of
2463 supplementary groups returned by the group upcall.
2465 Severity : enhancement
2467 Description: Allow masking D_WARNING, D_ERROR messages from console
2468 Details : Console messages can now be disabled via lnet.printk.
2472 Description: User code with malformed file open parameter crashes client node
2473 Details : Before packing join_file req, all the related reference should be
2474 checked carefully in case some malformed flags cause fake
2475 join_file req on client.
2479 Description: LDLM_ENQUEUE races with LDLM_CP_CALLBACK
2480 Details : ldlm_completion_ast() assumes that a lock is granted when the req
2481 mode is equal to the granted mode. However, it should also check
2482 that LDLM_FL_CP_REQD is not set.
2486 Description: Heavy nfs access might result in deadlocks
2487 Details : After ELC code landed, it is now improper to enqueue any mds
2488 locks under och_sem, because enqueue might want to decide to
2489 cancel open locks for same inode we are holding och_sem for.
2493 Description: Client eviction while running blogbench
2494 Details : A lot of unlink operations with concurrent I/O can lead to a
2495 deadlock causing evictions. To address the problem, the number of
2496 oustanding OST_DESTROY requests is now throttled to
2497 max_rpcs_in_flight per OSC and LDLM_FL_DISCARD_DATA blocking
2498 callbacks are processed in priority.
2502 Description: enable ACLs on MDS by default
2503 Details : ACLs must be enabled on MDS by default.
2506 Frequency : PPC/PPC64 only
2508 Description: conflicts between asm-ppc64/types.h and lustre_types.h
2509 Details : fix duplicated definitions between asm-ppc64/types.h and
2510 lustre_types.h on PPC.
2513 Frequency : PPC/PPC64 only
2515 Description: asm-ppc/segment.h does not exist
2516 Details : fix compile issue on PPC.
2520 Description: better handle error messages in extents code
2523 Frequency : RHEL4 only
2525 Description: mkfs is very slow on IA64/RHEL4
2526 Details : A performance regression has been discovered in the MPT Fusion
2527 driver between versions 3.02.73rh and 3.02.99.00rh. As a
2528 consequence, we have downgraded the MPT Fusion driver in the RHEL4
2529 kernel from 3.02.99.00 to 3.02.73 until this problem is fixed.
2531 Severity : enhancement
2533 Description: SNMP support enhancement
2534 Details : Adding total number of sampled request for an MDS node in snmp
2537 Severity : enhancement
2539 Description: Optimize ldlm waiting list processing for PR extent locks
2540 Details : When processing waiting list for read extent lock and meeting read
2541 lock that is same or wider to it that is not contended, skip
2542 processing rest of the list and immediatelly return current
2543 status of conflictness, since we are guaranteed there are no
2544 conflicting locks in the rest of the list.
2548 Description: Time out and refuse to reconnect
2549 Details : When the failover node is the primary node, it is possible
2550 to have two identical connections in imp_conn_list. We must
2551 compare not conn's pointers but NIDs, otherwise we can defeat
2552 connection throttling.
2556 Description: port llog fixes from b1_6 into HEAD
2557 Details : Port llog reference couting and some llog cleanups from b1_6
2558 (bug 10800) into HEAD, for protect from panic and access to already
2559 free llog structures.
2563 Description: Detect stride IO mode in read-ahead
2564 Details : When a client does stride read, read-ahead should detect that and
2565 read-ahead pages according to the detected stride pattern.
2569 Description: data checksumming impacts single node performance
2570 Details : add support for several checksum algorithm. Currently, only CRC32
2571 and Adler-32 are supported. The checksum type can be changed on
2572 the fly via /proc/fs/lustre/osc/*/checksum_type.
2576 Description: use adler32 for page checksums
2577 Details : when available, use the Adler-32 algorithm instead of CRC32 for
2582 Description: build for x2 fails
2583 Details : fix compile issue on Cray systems.
2587 Description: Properly match for duplicate locks
2588 Details : Due to different lock order from skiplists code, we need to
2589 traverse entire list for now
2592 Frequency : only on PPC/SLES10
2594 Description: "BITS_PER_LONG is not 32 or 64" in linux/idr.h
2595 Details : On SLES10/PPC, fs.h includes idr.h which requires BITS_PER_LONG to
2596 be defined. Add a hack in mkfs_lustre.c to work around this compile
2601 Description: LASSERT on MDS when client holding flock lock dies
2602 Details : ldlm pool logic depends on number of granted locks equal to
2603 number of released locks which is not true for flock locks, so
2604 just exclude such locks from consideration.
2608 Description: MDS deadlock with many ll_sync_lov threads and I/O stalled
2609 Details : Use fsfilt_sync() for both the whole filesystem sync and
2610 individual file sync to eliminate dangerous inode locking
2611 with I_LOCK that can lead to a deadlock.
2615 Description: performance in 1.6.3
2616 Details : Force q->max_phys_segments to MAX_PHYS_SEGMENTS on SLES10 to be
2617 sure that 1MB requests are not fragmented by the block layer.
2619 Severity : enhancement
2621 Description: organize the server-side client stats on per-nid basis
2622 Details : Change the structure of stats under obdfilter and mds to
2630 The "uuid"s file would list the uuids of _active_ exports.
2631 And the clear entry is to clear all stats and stale nids.
2633 Severity : enhancement
2635 Description: eliminate client locks in face of contention
2636 Details : file contention detection and lockless i/o implementation
2637 for contended files.
2641 Description: Reinitialize optind to 0 so that interactive lfs works in all cases
2644 Frequency : very rare, if additional xattrs are used on kernels >= 2.6.12
2646 Description: MDS may lose file striping (and hence file data) in some cases
2647 Details : If there are additional extended attributes stored on the MDS,
2648 in particular ACLs, SELinux, or user attributes (if user_xattr
2649 is specified for the client mount options) then there is a risk
2650 of attribute loss. Additionally, the Lustre file striping
2651 needs to be larger than default (e.g. striped over all OSTs),
2652 and an additional attribute must be stored initially in the
2653 inode and then increase in size enough to be moved to the
2654 external attribute block (e.g. ACL growing in size) for file
2659 Description: skiplist implementation simplification
2660 Details : skiplists are used to group compatible locks on granted list
2661 that was implemented as tracking first and last lock of each
2662 lock group the patch changes that to using doubly linked lists
2666 Description: MDS LBUG: ASSERTION(!IS_ERR(dchild))
2667 Details : Change LASSERTs to client eviction (i.e. abort client's recovery)
2668 because LASSERT on both the data supplied by a client, and the
2669 data on disk is dangerous and incorrect.
2671 Severity : enhancement
2673 Description: Slow truncate/writes to huge files at high offsets.
2674 Details : Directly associate cached pages to lock that protect those pages,
2675 this allows us to quickly find what pages to write and remove
2676 once lock callback is received.
2680 Description: more ldlm soft lockups
2681 Details : In ldlm_resource_add_lock(), call to ldlm_resource_dump()
2682 starve other threads from the resource lock for a long time in
2683 case of long waiting queue, so change the debug level from
2684 D_OTHER to the less frequently used D_INFO.
2686 Severity : enhancement
2688 Description: add -gid, -group, -uid, -user options to lfs find
2692 Description: Hung threads in invalidate_inode_pages2_range
2693 Details : The direct IO path doesn't call check_rpcs to submit a new RPC
2694 once one is completed. As a result, some RPCs are stuck in the
2695 queue and are never sent.
2699 Description: filter threads hungs on waiting journal commit
2700 Details : Cleanup filter group llog code, then only filter group llog will
2701 be only created in the MDS/OST syncing process.
2705 Description: Procfs and llog threads access destoryed import sometimes.
2706 Details : Sync the import destoryed process with procfs and llog threads by
2707 the import refcount and semaphore.
2709 Severity : enhancement
2711 Description: openlock cache of b1_6 port to HEAD
2716 Description: kernel BUG at ldiskfs2_ext_new_extent_cb
2717 Details : If insertion of an extent fails, then discard the inode
2718 preallocation and free data blocks else it can lead to duplicate
2723 Description: don't always update ctime in ext3_xattr_set_handle()
2724 Details : Current xattr code updates inode ctime in ext3_xattr_set_handle.
2725 In some cases the ctime should not be updated, for example for
2726 2.0->1.8 compatibility it is necessary to delete an xattr and it
2727 should not update the ctime.
2731 Bugzilla : 15713/16362
2732 Description: Assertion in iopen_connect_dentry in 1.6.3
2733 Details : looking up an inode via iopen with the wrong generation number can
2734 populate the dcache with a disconneced dentry while the inode
2735 number is in the process of being reallocated. This causes an
2736 assertion failure in iopen since the inode's dentry list contains
2737 both a connected and disconnected dentry.
2741 Description: assertion failure in ldlm_handle2lock()
2742 Details : fix a race between class_handle_unhash() and class_handle2object()
2743 introduced in lustre 1.6.5 by bug 13622.
2748 Description: Kernel BUG: sd_iostats_bump: unexpected disk index
2749 Details : remove the limit of 256 scsi disks in the sd_iostat patch
2754 Description: oops in sd_iostats_seq_show()
2755 Details : unloading/reloading the scsi low level driver triggers a kernel
2756 bug when trying to access the sd iostat file.
2761 Description: Kernel panics during QLogic driver reload
2762 Details : REQ_BLOCK_PC requests are not handled properly in the sd iostat
2763 patch, causing memory corruption.
2768 Description: journal_dev option does not work in b1_6
2769 Details : pass mount option during pre-mount.
2771 Severity : enhancement
2773 Description: Add a FIEMAP(FIle Extent MAP) ioctl
2774 Details : FIEMAP ioctl will allow an application to efficiently fetch the
2775 extent information of a file. It can be used to map logical blocks
2776 in a file to physical blocks in the block device.
2780 Description: LDLM soft lockups - improvement
2781 Details : It is be possible to send the lock handle along with each read
2782 or write request because the client is already doing a lock match
2783 itself so there isn't any reason the OST should have to re-do that
2787 Frequency : only on Cray X2
2789 Description: X2 build failures
2790 Details : fix build failures on Cray X2.
2794 Description: xid & resent requests
2795 Details : Initialize RPC XID from clock at startup (randomly if clock is
2798 Severity : enhancement
2800 Description: Add lustre_start utility to start or stop multiple Lustre servers
2805 Description: Lustre GPF in {:ptlrpc:ptlrpc_server_free_request+373}
2806 Details : In case of memory pressure, list_del() can be called twice on
2807 req->rq_history_list, causing a kernel oops.
2811 Description: kptllnd_peer_check_sends()) ASSERTION(!in_interrupt()) failed
2812 Details : fix stack overflow in the distributed lock manager by defering
2813 export eviction after a failed AST to the elt thread instead of
2814 handling it in the dlm interpret routine.
2818 Description: Convert some comments to new format.
2819 Details : Update documenting comments to match doxygen conventions.
2823 Description: Grammar fixes.
2824 Details : A couple of trivial spelling fixes.
2828 Description: OSD_COUNTERS-mandatory
2829 Details : Make previously optional ->oti_{w,r}_locks sanity checks mandatory
2830 to simplify the code.
2834 Description: simplify cmm_device freeing logic.
2835 Details : Call cmm_device_free() in the failure path of cmm_device_alloc().
2839 Description: Add lockdep support to dt_object_operations locking interface.
2840 Details : Augment ->do_{read,write}_lock() prototypes with a `role'
2841 parameter indicating lock ordering. Update mdd code to use new
2846 Description: Introduce failloc constants for lockless IO tests.
2847 Details : Add two new failloc constants to test lockless IO. Only one of
2848 them in implemented---another is checked in yet to be landed
2853 Description: Add lockdep support for inode mutex.
2854 Details : Introduce and use new LOCK_INODE_MUTEX_PARENT() macro to be used
2855 in the situations where Lustre has to lock more than one inode
2860 Description: Add optional invariants checking support.
2861 Details : Add new LINVRNT() macro, optional on new --enable-invariants
2862 configure switch. This macro is to be used for consistency and
2863 sanity checks that are too expensive to be left in `production'
2868 Description: Zap lock->l_granted_mode with explicit LCK_MINMODE.
2869 Details : Use LCK_MINMODE rather than 0 to reset lock->l_granted_mode to
2874 Description: Add lockdep support for ldlm_lock and ldlm_resource.
2875 Details : Use spin_lock_nested() in (the only) situation where more than
2876 one ldlm_lock is locked simultaneously. Also, fix possible
2877 dead-lock in ldlm_lock_change_resource() by enforcing particular
2882 Description: Use struct ldlm_callback_suite in ldlm_lock_create().
2883 Details : Instead of specifying each ldlm_lock call-back through separate
2884 parameter, wrap them into struct ldlm_callback_suite.
2888 Description: Kill join_lru obd method and its callers.
2889 Details : CLIO uses lock weighting policy to keep locks over mmapped regions
2890 in memory---a requirement implemented through ->o_join_lru() obd
2891 method in HEAD. Remove this method and its users.
2895 Description: Add asynchronous ldlm ENQUEUE completion handler.
2896 Details : CLIO posts ENQUEUE requests asynchronously through ptlrpcd---a
2897 case that stock ldlm_completion_ast() cannot handle as it waits
2898 until lock is granted. Introduce new ldlm_completion_ast_async()
2899 for this. Also comment ldlm_completion_ast().
2903 Description: ldlm_error <-> errno conversion.
2904 Details : Add functions to map (rather arbitrary) between LDLM error codes
2905 and standard errno values. CLIO needs this to prevent LDLM specific
2906 constants from escaping ldlm and osc.
2910 Description: Kill unused ldlm_handle2lock_ns() function.
2911 Details : Kill unused ldlm_handle2lock_ns() function.
2915 Description: Add lu_ref support to ldlm_lock
2916 Details : lu_ref support for ldlm_lock and ldlm_resource. See lu_ref patch.
2917 lu_ref fields ->l_reference and ->lr_reference are added to ldlm_lock
2918 and ldlm_resource. LDLM interface has to be changed, because code that
2919 releases a reference on a lock, has to "know" what reference this is.
2920 In the most frequent case
2922 lock = ldlm_handle2lock(handle);
2924 LDLM_LOCK_PUT(lock);
2926 no changes are required. When any other reference (received _not_ from
2927 ldlm_handle2lock()) is released, LDLM_LOCK_RELEASE() has to be called
2928 instead of LDLM_LOCK_PUT().
2930 Arguably, changes are pervasive, and interface requires some discipline
2931 for proper use. On the other hand, it was very instrumental in finding
2932 a few leaked lock references.
2936 Description: Add ldlm_lock_addref_try().
2937 Details : Introduce ldlm_lock_addref_try() function (used by CLIO) that
2938 attempts to addref a lock that might be being canceled
2943 Description: Add ldlm_weigh_callback().
2944 Details : Add new ->l_weigh_ast() call-back to ldlm_lock. It is called
2945 by ldlm_cancel_shrink_policy() to estimate lock "value", instead of
2946 hard-coded `number of pages' logic.
2950 Description: Add lockdep annotations to llog code.
2951 Details : Use appropriately tagged _nested() locking calls in the places
2952 where llog takes more than one ->lgh_lock lock.
2956 Description: Add loi_kms_set().
2957 Details : Wrap kms updates into a helper function.
2961 Description: Constify instances of struct lsm_operations.
2962 Details : Constify instances of struct lsm_operations.
2966 Description: lu_conf support.
2967 Details : On a server, a file system object is uniquely identified
2968 by a fid, which is sufficient to locate and load all object
2969 state (inode). On a client, on the other hand, more data are
2970 necessary instantiate an object. Change lu_object_find() and
2971 friends to take additional `lu_conf' argument describing object.
2972 Typically this includes layout information.
2976 Description: lu_context fixes.
2977 Details : Introduce new lu_context functions that are needed on the client
2978 side, where some system threads (ptlrpcd) are shared by multiple
2979 modules, and so cannot be stopped during module shutdown.
2983 Description: Add start and stop methods to lu_device_type_operations.
2984 Details : Introduce two new methods in lu_device_type_operations, that are
2985 invoked when first instance of a given type is created and last one
2986 is destroyed respectively. This is need by CLIO.
2990 Description: Add lu_ref support to struct lu_device.
2991 Details : Add lu_ref support to lu_object and lu_device. lu_ref is used to
2992 track leaked references.
2996 Description: Introduce lu_kmem_descr.
2997 Details : lu_kmem_descr and its companion interface allow to create
2998 and destroy a number of kmem caches at once.
3002 Description: Fix lu_object finalization race.
3003 Details : Fix a race between lu_object_find() finding an object and its
3004 concurrent finalization. This race is (most likely) not possible
3005 on the server, but might happen on the client.
3009 Description: Introduce lu_ref interface.
3010 Details : lu_ref is a debugging module allowing to track references to
3011 a given object. It is quite cpu expensive, and has to be
3012 explicitly enabled with --enable-lu_ref. See usage description
3017 Description: Factor lu_site procfs stats into a separate function.
3018 Details : Separate lu_site stats printing code into a separate function
3019 that can be reused on a client.
3023 Description: Constify instances of struct {lu,dt,md}_device_operations.
3024 Details : Constify instances of struct {lu,dt,md}_device_operations.
3028 Description: Introduce struct md_site and move meta-data specific parts of
3029 struct lu_site here.
3030 Details : Move md-specific fields out of struct lu_site into special struct
3031 md_site, so that lu_site can be used on a client.
3035 Description: Kill mdd_lov_destroy().
3036 Details : Remove unused mdd code.
3040 Description: Add st_block checking to multistat.c.
3041 Details : Add st_block checking to multistat.c.
3045 Description: Add lu_ref support to struct obd_device.
3046 Details : Add lu_ref tracking to obd_device.
3050 Description: Kill obd_set_fail_loc().
3051 Details : Remove unused code.
3055 Description: Add special type for ptlrpc_request interpret functions.
3056 Details : Add lu_env parameter to ->rq_interpreter call-back. NULL is passed
3057 there. Actual usage will be in CLIO.
3061 Description: Replace RW_LOCK_UNLOCKED() macro with rwlock_init().
3062 Details : Replace RW_LOCK_UNLOCKED() with rwlock_init() as the former
3063 doesn't work with lockdep.
3067 Description: Add rwv.c test program.
3068 Details : New testing program exercising readv(2) and writev(2) (Qian).
3072 Description: Add sendfile.c test program.
3073 Details : New testing program exercising sendfile(2) (Jay).
3077 Description: Ratelimit a message that can be very frequent.
3078 Details : Ratelimit a memory allocation failure message that can
3083 Description: Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
3084 Details : Use cdebug_show() in CDEBUG-style macros defined outside of libcfs.
3088 Description: Liblustre build fixes.
3089 Details : Liblustre build fixes.
3093 Description: libcfs: add cfs_{need,cond}_resched() interface.
3094 Details : libcfs: add cfs_{need,cond}_resched() definition and
3095 implementations for Linux, NT, and liblustre.
3097 Severity : enhancement
3099 Description: More exported tunables for mballoc
3100 Details : Add support for tunable preallocation window and new tunables for
3101 large/small requests
3105 Description: Detect corruption of block bitmap and checking for preallocations
3106 Details : Checks validity of on-disk block bitmap. Also it does better
3107 checking of number of applied preallocations. When corruption is
3108 found, it turns filesystem readonly to prevent further corruptions.
3112 Description: (rw.c:1323:ll_read_ahead_pages()) ASSERTION(page_idx > ria->ria_stoff) failed
3113 Details : Once the unmatched stride IO mode is detected, shrink the stride-ahead
3114 window to 0. If it does hit cache miss, and read-pattern is still
3115 stride-io mode, does not reset the stride window, but also does not
3116 increase the stride window length in this case.
3120 Frequency : only for big-endian servers
3121 Description: Check if system is big-endian while mounting fs with extents feature
3122 Details : Mounting a filesystem with extents feature will fail on big-endian
3123 systems since ext3-based ldiskfs is not supported on big-endian
3124 systems. This can be over-riden with "bigendian_extents" mount option.
3126 Severity : enhancement
3128 Description: The root squash functionality
3129 Details : A security feature, which is to prevent users from being able
3130 to mount lustre on their desktop, run as root, and delete
3131 all of the files in the filesystem. The goal is accomplished by
3132 remapping user id (UID) and group id (GID) of the root user to
3133 a UID and GID specified by the system administartor via Lustre
3134 configuration management server (MGS). The functionality also
3135 allows to specify sets of clients for which the remapping does
3140 Description: Excessive recovery window
3141 Details : With AT enabled, the recovery window can be excessively long (6000+
3142 seconds). To address this problem, we no longer use
3143 OBD_RECOVERY_FACTOR when extending the recovery window (the connect
3144 timeout no longer depends on the service time, it is set to
3145 INITIAL_CONNECT_TIMEOUT now) and clients report the old service
3146 time via pb_service_time.
3150 Description: Watchdog triggered on MDS failover
3151 Details : enable OBD_CONNECT_MDT flag when connecting from the MDS so that
3152 the OSTs know that the MDS "UUID" can be reused for the same export
3153 from a different NID, so we do not need to wait for the export to be
3157 Frequency : rare, only if using MMP with Linux RAID
3159 Description: MMP doesn't work with Linux RAID
3160 Details : While using HA for Lustre servers with Linux RAID, it is possible
3161 that MMP will not detect multiple mounts. To make this work we
3162 need to unplug the device queue in RAID when the MMP block is being
3163 written. Also while reading the MMP block, we should read it from
3164 disk and not the cached one.
3166 Severity : enhancement
3168 Description: open file using fid
3169 Details : A file can be opened using just its fid, like
3170 <mntpt>/.lustre/fid/SEQ:OID:VER - this is needed for HSM and replication
3173 Frequency : Only in RHEL5 when mounting multiple ext3 filesystems
3176 Description: "kmem_cache_create: duplicate cache jbd_4k" error message
3177 Details : add proper locking for creation of jbd_4k slab cache
3181 Description: MMP check in ext3_remount() fails without displaying any error
3182 Details : When multiple mount protection fails during remount, proper error
3185 Severity : enhancement
3187 Description: Allow stripe size to be up to 4G-64k
3188 Details : Fix math logic to allow large stripe sizes.
3192 Description: add check for >8TB ldiskfs filesystems
3193 Details : ext3-based ldiskfs does not support greater than 8TB LUNs.
3194 Don't allow >8TB ldiskfs filesystems to be mounted without
3195 force_over_8tb mount option
3200 Descriptoin: OSS complains "too many blocks" and crashes
3201 Details : In filter_commitrw_write():
3202 1. lquota_chkquota();
3203 2. obdo_from_inode();
3204 3. lquota_pending_commit();
3205 "2." has changed uid and gid in this case, then caused
3206 lquota_chkquota() and lquota_pending_commit() to use
3207 different uid and gid. That is the root of the bug.
3211 Descriptoin: getstripe doesn't report stripe count = -1 sanely
3212 Details : Detect and print out the 'stripe count = -1' case properly.
3216 Description: Out or order replies might be lost on replay
3217 Details : In ptlrpc_retain_replayable_request if we cannot find retained
3218 request with tid smaller than one currently being added, add it
3219 to the start, not end of the list.
3221 --------------------------------------------------------------------------------
3223 2007-08-10 Cluster File Systems, Inc. <info@clusterfs.com>
3225 * Support for kernels:
3226 2.6.5-7.283 (SLES 9),
3227 2.6.9-55.EL (RHEL 4),
3228 2.6.16.46-0.14 (SLES 10),
3229 2.6.18.8 vanilla (kernel.org)
3230 * Client support for unpatched kernels:
3231 (see http://wiki.lustre.org/index.php?title=Patchless_Client)
3232 2.6.16 - 2.6.22 vanilla (kernel.org)
3233 * Due to recently discovered recovery problems, we do not recommend
3234 using patchless RHEL 4 clients with this or any earlier release.
3235 * Recommended e2fsprogs version: 1.39.cfs8
3236 * Note that reiserfs quotas are disabled on SLES 10 in this kernel.
3237 * Starting with this release, the ldiskfs backing filesystem required
3238 by Lustre is now in its own package, lustre-ldiskfs. This package
3239 should be installed. It is versioned separately from Lustre and
3240 may be released separately in future.
3244 Description: block reactivating mgc import until all deactivates complete
3245 Details : Fix race when failing back MDT/MGS to itself (testing)
3248 Frequency : at statup only
3250 Description: mds_lov_synchronize race leads to various problems
3251 Details : simultaneous MDT->OST connections at startup can cause the
3252 sync to abort, leaving the OSC in a bad state.
3254 Severity : enhancement
3256 Description: add optional extra BUILD_VERSION info
3257 Details : add a new environment variable (namely LUSTRE_VERS) which allows
3258 to override the lustre version.
3261 Frequency : 2.6.18 servers only
3263 Description: ll_kern_mount() doesn't release the module reference
3264 Details : The ldiskfs module reference count never drops down to 0
3265 because ll_kern_mount() doesn't release the module reference.
3270 Description: server LBUG when using old ost_num_threads parameter
3271 Details : Accept the old ost_num_threads parameter but warn that it
3272 is deprecated, and fix an off-by-one error that caused an LBUG.
3277 Description: Transient SCSI error results in persistent IO issue
3278 Details : iobuf->dr_error is not reinitialized to 0 between two
3282 Frequency : sometimes when underlying device returns I/O errors
3284 Description: OSTs not going read-only during write failures
3285 Details : OSTs are not remounted read-only when the journal commit threads
3286 get I/O errors because fsfilt_ext3 calls journal_start/stop()
3287 instead of the ext3 wrappers.
3291 Description: poor connect scaling with increasing client count
3292 Details : Don't run filter_grant_sanity_check for more than 100 exports
3293 to improve scaling for large numbers of clients.
3296 Frequency : SLES10 only
3298 Description: sanity-quota.sh quotacheck failed: rc = -22
3299 Details : Quotas cannot be enabled on SLES10.
3302 Frequency : liblustre clients only
3304 Description: getdirentries does not give error when run on compute nodes
3305 Details : getdirentries does not fail when the size specified as an argument
3306 is too small to contain at least one entry
3308 Severity : enhancement
3310 Description: Add LNET router traceability for debug purposes
3311 Details : If a checksum failure occurs with a router as part of the
3312 IO path, the NID of the last router that forwarded the bulk data
3313 is printed so it can be identified.
3318 Description: OST "spontaneously" evicts client; client has imp_pingable == 0
3319 Details : Due to a race condition, liblustre clients were occasionally
3320 evicted incorrectly.
3322 Severity : enhancement
3324 Description: lfs setstripe use optional parameters instead of postional
3327 Severity : enhancement
3329 Description: Nanosecond timestamp support for ldiskfs
3330 Details : The on-disk ldiskfs filesystem has added support for nanosecond
3331 resolution timestamps. There is not yet support for this at
3332 the Lustre filesystem level.
3335 Frequency : during server recovery
3337 Description: MDS failing to send precreate requests due to OSCC_FLAG_RECOVERING
3338 Details : request with rq_no_resend flag not awake l_wait_event if they get
3342 Frequency : nfs export on patchless client
3344 Description: connectathon hang when test nfs export over patchless client
3345 Details : Disconnected dentry cannot be found with lookup, so we do not need
3346 to unhash it or make it invalid
3349 Description: fix llapi_lov_get_uuids() to allow many OSTs to be returned
3350 Details: : Change llapi_lov_get_uuids() to read the UUIDs from /proc instead
3351 of using an ioctl. This allows lfsck for > 160 OSTs to succeed.
3356 Description: open req refcounting wrong on reconnect
3357 Details : If reconnect happened between getting open reply from server and
3358 call to mdc_set_replay_data in ll_file_open, we will schedule
3359 replay for unreferenced request that we are about to free.
3360 Subsequent close will crash in variety of ways.
3361 Check that request is still eligible for replay in
3362 mdc_set_replay_data().
3367 Description: disable writes to filesystem when reading health_check file
3368 Details : the default for reading the health_check proc file has changed
3369 to NOT do a journal transaction and write to disk, because this
3370 can cause reads of the /proc file to hang and block HA state
3371 checking on a healthy but otherwise heavily loaded system. It
3372 is possible to return to the previous behaviour during configure
3373 with --enable-health-write.
3375 Severity : enhancement
3377 Description: 64-bit inode version
3378 Details: : Add a on-disk 64-bit inode version for ext3 to track changes made
3379 to the inode. This will be required for version-based recovery.
3384 Description: MDS fails to start if a duplicate client export is detected
3385 Details : in some rare cases it was possible for a client to connect to
3386 an MDS multiple times. Upon recovery the MDS would detect this
3387 and fail during startup. Handle this more gracefully.
3389 Severity : enhancement
3391 Description: Add -o localflock option to simulate old noflock
3393 Details : This will achieve local-only flock/fcntl locks
3399 Description: log_commit_thread vs filter_destroy race leads to crash
3400 Details : Take import reference before releasing llog record semaphore
3405 Description: Wrong request locking in request set processing
3406 Details : ptlrpc_check_set wrongly uses req->rq_lock for proctect add to
3407 imp_delayed_list, in this place should be used imp_lock.
3410 Frequency : when reconnection
3412 Description: Grant Leak when osc reconnect to OST
3413 Details : When osc reconnect ost, OST(filter) should check whether it
3414 should grant more space to client by comparing fed_grant and
3415 cl_avail_grant, and return the granted space to client instead
3416 of "new granted" space, because client will call osc_init_grant
3417 to update the client grant space info.
3420 Frequency : when client reconnect to OST
3422 Description: Grant Leak when osc do resend and replay bulk write
3423 Details : When osc reconnect to OST, OST(filter)should clear grant info of
3424 bulk write request, because the grant info while be sync between
3425 OSC and OST when reconnect, and we should ignore the grant info
3426 these of resend/replay write req.
3431 Description: Grant space more than available left space sometimes.
3432 Details : When then OST is about to be full, if two bulk writing from
3433 different clients came to OST. Accord the avaliable space of the
3434 OST, the first req should be permitted, and the second one
3435 should be denied by ENOSPC. But if the seconde arrived before
3436 the first one is committed. The OST might wrongly permit second
3437 writing, which will cause grant space > available space.
3440 Frequency : when client is evicted
3442 Description: Grant might be wrongly erased when osc is evicted by OST
3443 Details : when the import is evicted by server, it will fork another
3444 thread ptlrpc_invalidate_import_thread to invalidate the
3445 import, where the grant will be set to 0. While the original
3446 thread will update the grant it got when connecting. So if
3447 the former happened latter, the grant will be wrongly errased
3448 because of this race.
3453 Description: Checking Stale with correct fid
3454 Details : ll_revalidate_it should uses de_inode instead of op_data.fid2
3455 to check whether it is stale, because sometimes, we want the
3456 enqueue happened anyway, and op_data.fid2 will not be initialized.
3458 Severity : enhancement
3460 Description: update patchless client
3461 Details : Add support for patchless client with 2.6.20, 2.6.21 and RHEL 5
3464 Frequency : only with 2.4 kernel
3466 Description: random memory corruption
3467 Details : size of struct ll_inode_info is to big for union inode.u and this
3468 can be cause of random memory corruption.
3473 Description: Memory leak in recovery
3474 Details : Lov_mds_md was not free in an error handler in mds_create_object.
3475 It should also check obd_fail before fsfilt_start, otherwise if
3476 fsfilt_start return -EROFS,(failover mds during mds recovery).
3477 then the req will return with repmsg->transno = 0 and rc = EROFS.
3478 and we met hit the assert LASSERT(req->rq_reqmsg->transno ==
3479 req->rq_repmsg->transno) in ptlrpc_replay_interpret. Fcc should
3480 be freed no matter whether fsfilt_commit success or not.
3483 Frequency : only with huge count clients
3485 Description: Prevents from taking the superblock lock in llap_from_page for
3487 Details : using LL_ORIGIN_REMOVEPAGE origin flag instead of LL_ORIGIN_UNKNOW
3488 for llap_from_page call in ll_removepage prevents from taking the
3489 superblock lock for a soon died page.
3494 Description: Not check open intent error before release open handle
3495 Details : in some rare cases, the open intent error is not checked before
3496 release open handle, which may cause
3497 ASSERTION(open_req->rq_transno != 0), because it tries to release
3498 the failed open handle.
3503 Description: Set cat log bitmap only after create log success.
3504 Details : in some rare cases, the cat log bitmap is set too early. and it
3505 should be set only after create log success.
3509 Description: Accessing a block bevice can re-enable I/O when Lustre is
3510 tearing down a device.
3511 Details : dev_clear_rdonly(bdev) must be called in kill_bdev() ins