4 - limit Lustre IOVs to PTL_MD_MAX_IOV (611336)
6 2002-10-23 Phil Schwan <phil@clusterfs.com>
9 - in-use dentries weren't being reused properly (617851)
10 - prevent multiple LDLM setup (599178)
11 - fix LOV size calculations for truncate (617853)
12 - fix client handling of MDS intent errors (POSIX)
13 - fix permission bug in lovstripe.c test (624321)
14 - fix MDS thread deadlock (625069)
15 - truncate past end of file could corrupt data
16 - proper cleanup after timeouts, crashes, etc (592524, 550815)
17 - a race in recovery could return ETIMEDOUT to apps (623947)
18 - building outside the source directory was fixed
19 * the howto was brought (more) up to date (582544)
20 * major progress was made on recovery functionality
22 2002-10-10 Phil Schwan <phil@clusterfs.com>
25 - recovery deadlock fix
26 - rm -rf causes LBUG fix (617817)
27 - file open by multiple tasks fix (618962)
28 - directory permissions bugs (602707 and 620007)
29 - journal_stop fixed with locking (611313)
30 - O_APPEND failures resolved (618273, perhaps 614459)
31 - lconf PATH fix (619770)
32 - IA64 build fix (621450)
33 - RPC buffer sizes scale with amount of memory
35 2002-10-01 Phil Schwan <phil@clusterfs.com>
38 - locks would be cancelled without throwing away data pages,
39 resulting in inconsistent data (605627)
40 - inode attributes were not always being refreshed (605627, 612449)
41 - lconf now continues to cleanup after lctl reports an error
42 - MDS now enforces user permissions (602707)
43 - lprocfs cleanup fixed, but not yet enabled (614157)
44 - fixed infinite server hang, should a client not respond to an AST
45 - avoid going into recovery if user calls readlink() with a buffer
46 that's too small (613941)
47 - AST RPCs no longer require replies (614867) -- this may be changed
48 - don't crash server if client sends an IOV that's too big (611336)
49 - fixed lock conversion deadlock (611892)
50 - fixed the following of symlinks (614622)
51 * recovery: the server can remove locks from a client that dies, other
52 clients can make progress
53 * more extN patch fixes
54 * compile-time configurable ptlrpc buffer allocations
56 - collaborative read cache document
57 - Lustre Lite Performance CDR document-in-progress
59 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
62 - fix typo in patch-2.4.18
64 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
67 - clear ptlrpc request each time in handle_incoming_request()
68 - unlink of files now destroys the object on the OST
70 2002-09-19 Peter Braam <braam@clusterfs.com>
72 * add hard link support
73 * change obdfile creation method
74 * kernel patch changed
76 2002-09-19 Peter Braam <braam@clusterfs.com>
79 - stack overflow bug in extN fixed
81 2002-09-18 Andreas Dilger <adilger@clusterfs.com>
83 * documentation updates
84 - add man pages for config tools
85 - update tests/README to describe testing with new config tools
86 - finish metadata API descriptions
87 * bug fixes and cleanups
88 - statfs workaround for 16TB limit
89 - LOV stripe allocation improved, can stripe on subset of OSTs
90 - LOV file size/IO offset was wrong for files > 4GB in size
91 - object EA data was being dropped, caused files to be unreadable
92 - memory overflow with non-LOV OST caused memory corruption
93 - fixed regression tests to work with new config tools, obdfilter
94 - fixed bug when directory size became larger than 1 block
95 - fixed bug (for single client case) when PWD was deleted
96 - invalidate local directory pages when doing intent-based ops
97 - avoid LDLM oops when lock callback contained bad data
99 2002-09-09 Andreas Dilger <adilger@clusterfs.com>
101 * documentation updates
102 * bug fixes and cleanups
103 - configuration tools
106 - 64-bit compile warnings
107 - 64-bit internal statfs data
109 * test_brw on persistent OST devices
113 2002-09-04 Andreas Dilger <adilger@clusterfs.com>
115 * documentation updates
116 * bug fixes and cleanups
117 * configuration tools
119 2002-08-30 Peter J. Braam <braam@clusterfs.com>
122 * many small fixes to 0_5_4
123 * io/network handling
124 * thinkos in MDS operations
126 2002-08-24 Peter J. Braam <braam@clusterfs.com>
129 * crucial basic fixes to 0.5.3
130 * IOR, Iozone work over Elan
133 2002-08-07 Phil Schwan <phil@clusterfs.com>
134 * version 0_5_3, our first alpha
135 * we use the new Portals iovs
136 * documentation updates
137 * bug fixes and cleanups
138 * small changes in the DLM wire protocol
140 2002-07-25 Peter J. Braam <braam@clusterfs.com>
141 * version 0_5_1 with some initial stability,
142 * locking on MD and file I/O.
143 * documentation updates
144 * several bug fixes since 0.5.0
145 * small changes in wire protocol
147 2002-07-18 Phil Schwan <phil@clusterfs.com>
149 * delivered as Lustre Light Alpha
150 * fixed a crash after handling invalid MDS requests
151 * fixed directory pages for architectures with non-4k pages sizes
153 2002-07-11 Andreas Dilger <adilger@clusterfs.com>
154 * release version v0_4_4
155 * Moves TCP acceptor to be on port 2432 (unused Coda port) instead
157 * Fixes a number of interruption problems with OST operations.
158 * Update documentation for portals header changes
159 * Move all wire protocol structs/defines to lustre_idl.h
160 * Fixes symlink length bug.
161 * Add tcpdump to repository.
163 2002-07-05 Andreas Dilger <adilger@clusterfs.com>
164 * release version v0_4_3
165 * Fixes statfs for inodes on extN.
166 * Fixes bug in runtests which would delete /etc/hosts.
167 * Use 64-bit object IDs wherever possible (not into VFS though)
168 Remove ost_get_info, which is unused by lustre, and out of date.
170 2002-07-03 Peter Braam <braam@clusterfs.com>
171 * release version v0_4_2 Fixes a lookup error (type not passed)
172 * move forward to head of Portals
173 * move forward to latest Lustre kernel
175 2002-06-25 Peter Braam <braam@clusterfs.com>
176 * release version v0_4_1. Hopefully stable on single node use.