sin/samba-mirror
1
0
Fork 0
mirror of https://github.com/samba-team/samba.git synced 2025-01-12 09:18:10 +03:00
Code
1,509 Commits 60 Branches 1,145 Tags 452 MiB
0e31eaed57
Commit Graph

230 Commits

Author SHA1 Message Date
Ronnie Sahlberg
3f56526037 Specify and print debuglevels by name and not by number 
(This used to be ctdb commit 79ad830294b8b677fbd0c5ad7ed6fbde71f74f8d)
 2008-02-05 10:26:23 +11:00
Andrew Tridgell
f6e53f433b merge from ronnie 
(This used to be ctdb commit e7b57d38cf7255be823a223cf15b7526285b4f1c)
 2008-02-04 20:07:15 +11:00
Andrew Tridgell
9d6ac0cf55 added debug constants to allow for better mapping to syslog levels 
(This used to be ctdb commit 7ba8f1dde318eab03f4257e5a89fd23e7281e502)
 2008-02-04 17:44:24 +11:00
Andrew Tridgell
feb7c05734 removed dependence on dprintf 
(This used to be ctdb commit c156db449218bf9432e3a6cb3ce0f617197c9069)
 2008-01-29 14:31:51 +11:00
Andrew Tridgell
146d4b0db7 merge async recovery changes from Ronnie 
(This used to be ctdb commit 576e317640d25f8059114f15c6f1ebcee5e5b6e2)
 2008-01-29 13:59:28 +11:00
Andrew Tridgell
eb044bb1d6 make ctdb dumpmemory work remotely, and dump the talloc 
memory tree to stdout. This is much more useful than putting it in the log, and also fixes
a bug where the pipe would overflow internally and cause ctdbd to lockup

(This used to be ctdb commit e236979e2162d9bd7a495086342168a696cf76c5)
 2008-01-22 14:22:41 +11:00
Andrew Tridgell
d945b1af03 merge from ronnie 
(This used to be ctdb commit 5f6d59b9d18c694d82591238bc7a6bb98726a3ed)
 2008-01-17 16:46:56 +11:00
Ronnie Sahlberg
9625483c2d add ctdb_uptime.c 
(This used to be ctdb commit 4c7153681ed4d68d601720d043f9ff95ac7647a9)
 2008-01-17 16:37:05 +11:00
Ronnie Sahlberg
9055978b46 add a ctdb uptime command that prints when ctdb was started and when the 
last recovery occured

(This used to be ctdb commit b86e8ccbdac044bb949c4fc2ebb27635126272a9)
 2008-01-17 11:33:23 +11:00
Andrew Tridgell
5683a8d1e1 cope better with large debug dumps 
(This used to be ctdb commit fc3733f8e966376f50799fd1aa7b0a8e1cf66e0e)
 2008-01-16 23:06:37 +11:00
Andrew Tridgell
be9594c156 fixed handling of \r from stdout of subprocesses 
(This used to be ctdb commit f1acec5db4948d8e48412a8546bb181b08a2c5fd)
 2008-01-16 22:40:01 +11:00
Andrew Tridgell
0080683da8 fixed two 64bit warnings 
(This used to be ctdb commit c61fe240713ae2e917f69f827c6927405f02f5d4)
 2008-01-16 22:16:15 +11:00
Andrew Tridgell
97ede94e40 The recovery daemon does not need to be a realtime task 
(This used to be ctdb commit f552acf7c1f9dd37eb35d9716ea3fb02304aae8f)
 2008-01-16 22:08:33 +11:00
Andrew Tridgell
b62b7fcde8 added syslog support, and use a pipe to catch logging from child processes to the ctdbd logging functions 
(This used to be ctdb commit 1306b04cd01e996fd1aa1159a9521f2ff7b06165)
 2008-01-16 22:03:01 +11:00
Ronnie Sahlberg
5b7838d768 ctdb_control_send() does not need to take an outdata parameter 
remove the outdata parameter from the function and all callers

(This used to be ctdb commit e3951337f8df2ae19cce61c954036590c7a03582)
 2008-01-16 10:23:26 +11:00
Andrew Tridgell
bf9e33d4cf - catch a case where the client disconnects during a call 
- track all talloc memory, using NULL context

(This used to be ctdb commit bf89c56002f5311520e91cb367753bc46e5dddc9)
 2008-01-16 09:44:48 +11:00
Andrew Tridgell
6c56e9d347 fixed a memory leak in the recovery daemon 
(This used to be ctdb commit 73c27cf4c62cbe44b2b8fd00f907974d0808500c)
 2008-01-15 20:11:44 +11:00
Ronnie Sahlberg
ba31feaec0 split node health monitoring and checking for connected/disconnected 
nodes into two separate files.

move the monitoring of keepalives for detecting connected/disconnected 
remote nodes into ctdb_keepalive.c

(This used to be ctdb commit 23a57b20c314d5f11a433cf251eb9d9de743849a)
 2008-01-15 08:42:12 +11:00
Andrew Tridgell
b866a147d2 get rid of monitor_retry as well 
(This used to be ctdb commit c957cf9c1d99d5d3f4ca726f7a867c829660a2b7)
 2008-01-10 14:49:43 +11:00
Andrew Tridgell
538f519dba exponential backoff in health monitoring for faster startup 
(This used to be ctdb commit 1b04a1f675f73b48366ba98803a58c3d8df1b6e1)
 2008-01-10 14:40:56 +11:00
Andrew Tridgell
3b3fceacbe block alarm signals during critical sections of vacuum 
(This used to be ctdb commit cfb14ae76f00f10d27b56c034b2247ab12d63065)
 2008-01-10 09:43:14 +11:00
Andrew Tridgell
59d69bb709 only match vacuum list if on the same database 
(This used to be ctdb commit 27e56955e93027534780cc7549ddb224670d82b6)
 2008-01-09 10:22:20 +11:00
Andrew Tridgell
9559249e15 ensure the main daemon doesn't use a blocking lock on the freelist 
(This used to be ctdb commit 73f8257906b09e6516f675883d8e7a3c455ad869)
 2008-01-08 22:31:48 +11:00
Andrew Tridgell
1c91398aef ensure the recovery daemon is not clagged up by vacuum calls 
(This used to be ctdb commit ff7e80e247bf5a86adda0ef850d901478449675b)
 2008-01-08 21:28:42 +11:00
Andrew Tridgell
96100fcae6 added two new ctdb commands: 
ctdb vacuum   : vacuums all the databases, deleting any zero length
                 ctdb records

 ctdb repack   : repacks all the databases, resulting in a perfectly
                 packed database with no freelist entries

(This used to be ctdb commit 3532119c84ab3247051ed6ba21ba3243ae2f6bf4)
 2008-01-08 17:23:27 +11:00
Andrew Tridgell
25bb60f112 show start/stop time of recovery on all nodes 
(This used to be ctdb commit 9f7662279c367eb3e8a58e6f4aeca521e6f1f1d0)
 2008-01-08 09:30:11 +11:00
Andrew Tridgell
37861932ce merge from ronnie 
(This used to be ctdb commit 0aa6e04438aa5ec727815689baa19544df042cf7)
 2008-01-07 16:17:22 +11:00
Andrew Tridgell
d38fbaa38b nicer onnode output 
(This used to be ctdb commit ac5c1e090d007bc2e3965589731620b87c0217fb)
 2008-01-07 14:31:13 +11:00
Andrew Tridgell
4258098e98 catch internal traversal errors 
(This used to be ctdb commit 8caa85ad71be5d20a8d6f0cb3d52aff6905657a4)
 2008-01-07 14:08:25 +11:00
Andrew Tridgell
528e4d7a2b more efficient traversal in pulldb control 
(This used to be ctdb commit fe614b10868e63b70e081b5bbfb74bf16fdf5716)
 2008-01-07 14:07:01 +11:00
Andrew Tridgell
748843a3c6 added paranoid transaction ids 
(This used to be ctdb commit afc1da53873cdbd31fcc8c6b22fae262e344cf6e)
 2008-01-06 13:24:55 +11:00
Andrew Tridgell
c08f2616cd new simpler and much faster recovery code based on tdb transactions 
(This used to be ctdb commit 9ef2268a1674b01f60c58fed72af8ac982fe77a3)
 2008-01-06 12:38:01 +11:00
Andrew Tridgell
4f5b717aa3 change default tunables to cope with larger dbs 
(This used to be ctdb commit d91a2d43d1f0562cc3a12e6e1e2767f75d888f72)
 2008-01-06 12:36:58 +11:00
Andrew Tridgell
108aafcdb2 non-persistent databases don't need sync transactions 
(This used to be ctdb commit 52fd86addd23e4d6e0af2c716bd83d19675b1f5a)
 2008-01-06 12:36:30 +11:00
Andrew Tridgell
9311f7fb7e fixed the bug that make "onnode N service ctdb start" hang 
(This used to be ctdb commit b50dcb16f30a60abce42f491f9b0aae7948b8206)
 2008-01-05 12:09:29 +11:00
Andrew Tridgell
e4aefbc66d a new tunable DatabaseMaxDead that enables the tdb max dead cache logic 
(This used to be ctdb commit 01c519c3658a8fcb9545b507b597e723658e4c4e)
 2008-01-05 09:36:53 +11:00
Andrew Tridgell
023a230d9c a useful hack for checking correct behaviour of recovery 
(This used to be ctdb commit d88b95a5407b53ead47ca0638ee60653ea3d3d07)
 2008-01-05 09:36:21 +11:00
Andrew Tridgell
f79dfd04c0 convert much of the recovery logic to be async and parallel across all nodes 
(This used to be ctdb commit 8b72a02bf1045d8befb342a4111ca1316889262e)
 2008-01-05 09:35:43 +11:00
Andrew Tridgell
9a625534c1 this fixes the non-dmaster bug that has plagued us for months 
(This used to be ctdb commit 2acf6c6201862debfca054a09262f75c066d2deb)
 2008-01-05 09:34:47 +11:00
Andrew Tridgell
fc21f78231 make some specific cases of the non-dmaster bug non-fatal 
(This used to be ctdb commit 7b516ab06c7ba7ffe9ecf3f76720df5360176b2c)
 2008-01-05 09:32:29 +11:00
Andrew Tridgell
e9987cf236 fixed a warning 
(This used to be ctdb commit f34d0f9351c1cda3327efb14e173f249f7854570)
 2008-01-05 09:30:49 +11:00
Andrew Tridgell
afc7275c16 fixed a warning 
(This used to be ctdb commit d6255438d63943736b24a7a6da190b6933379a61)
 2008-01-04 12:42:10 +11:00
Andrew Tridgell
2509821503 prevent a re-ban loop for single node clusters 
(This used to be ctdb commit b20a3369655bcba274c99091157ba7466994e848)
 2008-01-04 12:11:29 +11:00
Andrew Tridgell
41fb8e283b add randrec to Makefile 
(This used to be ctdb commit ded1f7903e8a6525ab1888e8c4f50c71fa23cc19)
 2008-01-04 09:19:06 +11:00
Andrew Tridgell
bb06e831a0 more optimisations to recovery 
(This used to be ctdb commit 9a41ad0a842cd4f3792d6e84b5c809b7ff6f342e)
 2008-01-02 22:44:46 +11:00
Andrew Tridgell
2a2f1e3d91 fixed segv on failed ctdb_ctrl_getnodemap 
(This used to be ctdb commit 5daf9a72f0e60a9af7cf32ae6d759be7d94857ec)
 2007-12-27 10:07:01 +11:00
Andrew Tridgell
6ef3bff4ed merge from ronnie 
(This used to be ctdb commit 072ef744951d3aa59dd8be70578b99b18c37d988)
 2007-12-04 15:20:40 +11:00
Andrew Tridgell
a55c3709ea make DeterministicIPs the default 
(This used to be ctdb commit e7d077e98a40a62dbd6bfd174f29afba7b5529ef)
 2007-12-04 15:18:27 +11:00
Ronnie Sahlberg
7cef33b40a rework banning/unbanning nodes 
ctdb_recoverd.c
Always handle banning/unbanning locally on the node that is being 
banned/unbanned instead of on the recovery master.
This means that if a ban request comes in to the recovery master for a 
remote node, we pass the request on to the remote node instead of 
setting up the ban and ban timeouts locally.

ctdb.c
send ban/unban requests to the node being banned/unbanned instead of to 
the recmaster

(This used to be ctdb commit 880dd9f5fd0b91e450da93e195cc5c62cb1dcd6e)
 2007-12-03 15:45:53 +11:00
Ronnie Sahlberg
64008e28bb for the banned status, we should allocate this structure as a child of 
the banned_nodes array and not the rec structure so that  ban_state is 
destroyed when the banned_nodes array gets destroyed
(and so that when this struct is destroyed, that any pending 
ctdb_ban_timeout events are also destroyed.)

othervise we may end up with multiple ban_timeout timed events going in 
parallell since we destroy/recreate the banned_nodes structure during 
election   but we never destroy/recreate the rec structure.

(This used to be ctdb commit fbd663d56a2a4421a5c0e541962c87e2e9c7cd82)
 2007-12-03 11:39:17 +11:00