1 2002-10-01 Phil Schwan <phil@clusterfs.com>
4 - locks would be cancelled without throwing away data pages,
5 resulting in inconsistent data (605627)
6 - inode attributes were not always being refreshed (605627, 612449)
7 - lconf now continues to cleanup after lctl reports an error
8 - MDS now enforces user permissions (602707)
9 - lprocfs cleanup fixed, but not yet enabled (614157)
10 - fixed infinite server hang, should a client not respond to an AST
11 - avoid going into recovery if user calls readlink() with a buffer
12 that's too small (613941)
13 - AST RPCs no longer require replies (614867) -- this may be changed
14 - don't crash server if client sends an IOV that's too big (611336)
15 - fixed lock conversion deadlock (611892)
16 - fixed the following of symlinks (614622)
17 * recovery: the server can remove locks from a client that dies, other
18 clients can make progress
19 * more extN patch fixes
21 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
24 - fix typo in patch-2.4.18
26 2002-09-20 Andreas Dilger <adilger@clusterfs.com>
29 - clear ptlrpc request each time in handle_incoming_request()
30 - unlink of files now destroys the object on the OST
32 2002-09-19 Peter Braam <braam@clusterfs.com>
34 * add hard link support
35 * change obdfile creation method
36 * kernel patch changed
38 2002-09-19 Peter Braam <braam@clusterfs.com>
41 - stack overflow bug in extN fixed
43 2002-09-18 Andreas Dilger <adilger@clusterfs.com>
45 * documentation updates
46 - add man pages for config tools
47 - update tests/README to describe testing with new config tools
48 - finish metadata API descriptions
49 * bug fixes and cleanups
50 - statfs workaround for 16TB limit
51 - LOV stripe allocation improved, can stripe on subset of OSTs
52 - LOV file size/IO offset was wrong for files > 4GB in size
53 - object EA data was being dropped, caused files to be unreadable
54 - memory overflow with non-LOV OST caused memory corruption
55 - fixed regression tests to work with new config tools, obdfilter
56 - fixed bug when directory size became larger than 1 block
57 - fixed bug (for single client case) when PWD was deleted
58 - invalidate local directory pages when doing intent-based ops
59 - avoid LDLM oops when lock callback contained bad data
61 2002-09-09 Andreas Dilger <adilger@clusterfs.com>
63 * documentation updates
64 * bug fixes and cleanups
68 - 64-bit compile warnings
69 - 64-bit internal statfs data
71 * test_brw on persistent OST devices
75 2002-09-04 Andreas Dilger <adilger@clusterfs.com>
77 * documentation updates
78 * bug fixes and cleanups
81 2002-08-30 Peter J. Braam <braam@clusterfs.com>
84 * many small fixes to 0_5_4
86 * thinkos in MDS operations
88 2002-08-24 Peter J. Braam <braam@clusterfs.com>
91 * crucial basic fixes to 0.5.3
92 * IOR, Iozone work over Elan
95 2002-08-07 Phil Schwan <phil@clusterfs.com>
96 * version 0_5_3, our first alpha
97 * we use the new Portals iovs
98 * documentation updates
99 * bug fixes and cleanups
100 * small changes in the DLM wire protocol
102 2002-07-25 Peter J. Braam <braam@clusterfs.com>
103 * version 0_5_1 with some initial stability,
104 * locking on MD and file I/O.
105 * documentation updates
106 * several bug fixes since 0.5.0
107 * small changes in wire protocol
109 2002-07-18 Phil Schwan <phil@clusterfs.com>
111 * delivered as Lustre Light Alpha
112 * fixed a crash after handling invalid MDS requests
113 * fixed directory pages for architectures with non-4k pages sizes
115 2002-07-11 Andreas Dilger <adilger@clusterfs.com>
116 * release version v0_4_4
117 * Moves TCP acceptor to be on port 2432 (unused Coda port) instead
119 * Fixes a number of interruption problems with OST operations.
120 * Update documentation for portals header changes
121 * Move all wire protocol structs/defines to lustre_idl.h
122 * Fixes symlink length bug.
123 * Add tcpdump to repository.
125 2002-07-05 Andreas Dilger <adilger@clusterfs.com>
126 * release version v0_4_3
127 * Fixes statfs for inodes on extN.
128 * Fixes bug in runtests which would delete /etc/hosts.
129 * Use 64-bit object IDs wherever possible (not into VFS though)
130 Remove ost_get_info, which is unused by lustre, and out of date.
132 2002-07-03 Peter Braam <braam@clusterfs.com>
133 * release version v0_4_2 Fixes a lookup error (type not passed)
134 * move forward to head of Portals
135 * move forward to latest Lustre kernel
137 2002-06-25 Peter Braam <braam@clusterfs.com>
138 * release version v0_4_1. Hopefully stable on single node use.