Yehuda Sadeh
c2e1fa50bf
ceph-daemons: works partially
2009-02-27 15:46:10 -08:00
Sage Weil
4fc829114c
vstart: some changes
...
no more crun wrapper; rename ceph_stop back to stop.sh
2009-02-27 15:44:28 -08:00
Sage Weil
295fc872ea
vstart: no more crun
2009-02-27 15:44:28 -08:00
Sage Weil
2b9781080d
mkcephfs
...
Still need to add optional mkbtrfs + mount stuff.
2009-02-27 15:11:49 -08:00
Sage Weil
b13663e86e
debian: update packages
2009-02-27 15:11:49 -08:00
Sage Weil
268bb2c136
makefile, todo
2009-02-27 15:11:48 -08:00
Sage Weil
528dd9b0ce
config: specify --monmap_file on command line
...
Still defaults to .ceph_monmap.
2009-02-27 15:11:48 -08:00
Sage Weil
1819950069
client: fix warning
2009-02-27 15:11:48 -08:00
Sage Weil
e3312b51c6
ceph-daemon: some comments
2009-02-27 15:11:48 -08:00
Yehuda Sadeh
06572bbfbe
ceph-daemons: no norestart any more
2009-02-27 14:33:15 -08:00
Yehuda Sadeh
ab773a73d9
ceph-daemons: read config for different entities
2009-02-27 14:30:09 -08:00
Yehuda Sadeh
1a883973c4
crun: replace $* with "$@"
2009-02-27 12:55:41 -08:00
Yehuda Sadeh
3e78cb85bb
ceph-daemons: replace $* with "$@"
2009-02-27 12:51:51 -08:00
Sage Weil
1aed41ca74
crun: simplify to just restart if coredump
2009-02-27 12:43:10 -08:00
Yehuda Sadeh
c4eef372fd
cephd: added "osd journal" option
2009-02-27 12:36:38 -08:00
Sage Weil
fb330f6168
ceph-daemons: add --allhosts option
...
By default only start hosts on the current host if 'host' is defined. If
--allhosts is specified, ssh into the remote host to start/stop as needed.
2009-02-27 11:38:22 -08:00
Sage Weil
84a1ae3861
sample startup.conf
2009-02-27 11:18:45 -08:00
Sage Weil
357e8bbab8
ceph-daemons: restructured (but incomplete) init.d script
...
Only start/stop items explicitly defined in startup.conf.
Start specific osds or mdss (e.g. 'start osd12')
No NUM_OSD type iteration.. that only makes sense in a testing/dev
environment.
Extract options for each daemon in both start AND stop path, so that we can
stop the right daemon... e.g. if it is on a remote host or some such thing.
Try to be smart about looking for things in . or /usr/bin, /etc/ceph, etc.
2009-02-27 11:03:23 -08:00
Sage Weil
c2070b4cd3
config: optionally maintain a pid file
...
Remove it when we shut down _only_ if it matches getpid().
2009-02-27 10:57:24 -08:00
Sage Weil
6b5c13b65a
cconf: ability to list section names
...
'./cconf -c ceph.conf -l foo' will list all section names that begin with
'foo'.
2009-02-27 10:23:21 -08:00
Sage Weil
250f42d53e
ceph_init: assume CEPH_BIN is dirname ceph_init
2009-02-27 09:20:37 -08:00
Sage Weil
1490b7ff78
ceph_init: default to /etc/ceph/startup.conf
2009-02-27 09:05:19 -08:00
Sage Weil
ab14b5a1e5
debian: Build-depends libedit-dev
2009-02-27 08:46:41 -08:00
Sage Weil
238400d986
logger: validate log type; remove unused mds.logger2
...
Catch any undefined columns early. Only real non-startup runtime danger
is sending events to the wrong logger, now.
2009-02-26 20:47:11 -08:00
Sage Weil
e2389c5316
debug: ensure a SIGHUP reopens log reasonably soon
...
Since nothing new may be written to the log for a while, check if it needs
to be reopened in the mds/osd/mon tick() method.
2009-02-26 20:27:56 -08:00
Sage Weil
843c598e8b
logger: fix again
...
bah
2009-02-26 18:34:42 -08:00
Sage Weil
48c61bafdf
config: only build an abspath if it's not already absolute
2009-02-26 18:33:40 -08:00
Sage Weil
40459d5c51
config: default to etc/ceph/ceph.conf
2009-02-26 17:36:25 -08:00
Sage Weil
37f77cee29
debian: include var/log/ceph in package
2009-02-26 17:23:11 -08:00
Sage Weil
4d2c546583
logger: cleanup; make feeble attempt to create log dir
2009-02-26 17:23:03 -08:00
Sage Weil
8667c264f2
config: prefix default paths with install prefix
2009-02-26 17:13:43 -08:00
Sage Weil
40bb4187f8
config: verify argument exists before dereferencing args[++i]
...
Avoid segfault on malformed command line args.
2009-02-26 15:05:08 -08:00
Sage Weil
78f81687ef
config: use sane default paths for logs; use .conf for *start.sh
...
Specify sane deafult paths for logs (/var/log/ceph/*). Use workingdir.conf
when starting up from dstart.sh or vstart.sh to use directories relative
to the current (presumably, build) directory.
2009-02-26 14:50:38 -08:00
Sage Weil
48533a5c2d
logger: fixes
2009-02-26 14:48:40 -08:00
Sage Weil
61dd86d97e
logger: identify fields with int constants instead of strings
...
This avoids the crufty hashes and so forth. The only downside is that
_all_ fields must be predeclared. That's probably a good thing anyway.
Logger will assert if the key is invalid or the name hasn't been
registered. Hopefully I haven't missed any.
2009-02-26 11:25:37 -08:00
Sage Weil
9156a56bc5
config: remove ms_hosts
2009-02-26 09:52:20 -08:00
Sage Weil
849d3f4ca6
logger: configurable output dir
2009-02-26 09:52:00 -08:00
Sage Weil
131a2d8020
osd: strip out hard-coded osd device paths
...
The correct device path should be specified on the command line.
2009-02-26 09:47:42 -08:00
Sage Weil
6ff39baa45
kclient: zero_user_segments doesn't appear until 2.6.25
2009-02-26 09:23:34 -08:00
Sage Weil
13c67f63ba
kclient: more sysfs cleanup, bug fixes
2009-02-25 15:47:41 -08:00
Sage Weil
180a3d919b
kclient: fix kbuild patch, import script
2009-02-25 15:07:38 -08:00
Sage Weil
a21385a229
kclient: move parens
2009-02-25 15:07:24 -08:00
Sage Weil
95675f50c2
kclient: changes for 2.6.29
2009-02-25 14:40:55 -08:00
Sage Weil
dc6cdbec0e
kclient: avoid simple_prepare_write()
...
No longer exported.
2009-02-25 14:39:08 -08:00
Sage Weil
0d23c0923e
kclient: separate Kconfig
2009-02-25 14:34:55 -08:00
Sage Weil
a69f197e44
todos
2009-02-25 13:07:23 -08:00
Sage Weil
e9f470e204
kclient: add module author
2009-02-25 12:49:48 -08:00
Sage Weil
6adae97e87
kclient: dup kobj_sysfs_ops from kobject.c
...
They're not exported, bleh.
2009-02-25 12:49:18 -08:00
Sage Weil
662afd5ea5
journal: initiate full commit when journal hits 50% full
...
We want to avoid filling up the journal, particularly since
_start_sync is a no-op with a journal enabled.
2009-02-25 12:38:32 -08:00
Sage Weil
9643b96035
filestore: don't initiate sync on _start_sync if there's a journal
...
Journal will commit immediately anyway.. don't slow the rest
down with a full commit to disk.
2009-02-25 12:37:28 -08:00
