| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
Drop the config_type RPC req field, use just a "subop" key in the param dict.
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2816 ([glusterfs-3.2.0qa14]: gsyncd monitor crashed)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2816
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2346 (Log message enhancements in GlusterFS - phase 1)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2346
|
|
|
|
|
|
|
|
| |
Signed-off-by: Gaurav <gaurav@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2818 (Permission problems with gluster NFS works with native FUSE)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2818
|
|
|
|
|
|
|
|
|
|
| |
already present from parent directories.
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2697 (Quota: add-brick creates the size go awkward, though it was perfect earlier)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2697
|
|
|
|
|
|
|
|
|
|
| |
children a directory has in a single go.
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2809 ([e89f2a1f5daf1513c533a47b90aec8f90702d09f]: glusterfsd crashed in quota_dirty_inode_readdir)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2809
|
|
|
|
|
|
|
|
|
|
| |
mainly in unlink and rename
Signed-off-by: Amar Tumballi <amar@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2801 (glusterfsd process crash while running rebalance on simple volume)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2801
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- log-file does not need to be a reserved option anymore as we
wouldn't overwrite the user setting
- reimplement glusterd_gsync_get_param_file(): all use cases
have now pre-configured values, so instead of putting together
the needed pathname by ourselves, we can just get it from
gsyncd config
- refactor it to a common backend with glusterd_get_canon_url()
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is made possible by gsyncd config templating, by which
session specific settings can be expressed in a generic form.
Benefits:
- in glusterd we use only generic (rx pattern based) config settings
so we don't violate users' freedom to make settigs for their sessions
- don't have to invoke gsyncd in excess to prepare a session
- we can pre-configure slave side too (relying on the templates and
gsyncd service auto-discovery)
- much leaner, much more expressive code
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
So, for example, a log file setting of
/var/log/${mastervol}/${eSlave}.log
will be substituted with the volume name of the master and the
canonicalized-escaped name of the slave for each master-slave pair.
As template expanders, beyond the various forms and derivatives
of master and slave, the following are also available:
- gsyncd tunables (set in command line or in config)
- for regexp sections, regexp group captures can be accessed
via "match<i>_<n>", where i=1,2 corresponds to the i-th peer-rx
in the section title and n=1,... to the n-th capture.
This will enable us to have a static configuration (not having to
add new entries on each gsyncd start).
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
|
|
|
| |
If glusterd is running, gsyncd will find it out and adopt related settings
automagically (even if started from not by glusterd)
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
|
|
| |
Also make getwd command impatient, having it to time out in 1 sec
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2785 (gsyncd logs on slave side go to /dev/null)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2785
|
|
|
|
|
|
|
|
|
|
| |
session we adjust does not exist
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2777 (geo-replication config setting should have effect on already running sessions)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2777
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2736 (gsyncd hangs if crash occurs in the non-main thread)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2736
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2791 (trying to set up geo-replication with a gluster slave on the LAN screws up geo-rep irrecoverably)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2791
|
|
|
|
|
|
|
|
|
|
|
| |
- we want to enforce the separation of the volume name and the url namespaces
- resolving simple ids as relative paths was lame anyway
Signed-off-by: Csaba Henk <csaba@lowlife.hu>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2786 (Having to prepend geo-replication master vol with colon spoils the UI)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2786
|
|
|
|
|
|
|
|
| |
Signed-off-by: Pranith Kumar K <pranithk@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2797 ([glusterfs-3.2.0qa13]: peer probe fails with gsync running)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2797
|
|
|
|
|
|
|
|
| |
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2797 ([glusterfs-3.2.0qa13]: peer probe fails with gsync running)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2797
|
|
|
|
|
|
|
|
|
|
| |
are active on this volume.
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2765 (geo-replication should have mercy on brick failure)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2765
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
it is the last link.
- since we are not having different contributions for different (parent, name)
pairs, but only for different parents, links present in the same directory
will have only one contribution representing all of them. Hence parents will
not be updated for each link call and parents account for just file-size
instead of (numlinks * filesize). However this patch will cause issues
for links present accross directories, as only one of their contributions
is subtracted when all of them removed (Here all the links are accounted
in their parents, thereby total size accounted is [numlinks * filesize]).
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2689 (Quota: xattrs getting corrupted)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2689
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Though we are setting in memory contribution to zero, during updation
process contribution is actually fetched from backend. If they are not
deleted and if newpath did not exist before doing rename, contributions
from oldloc->inode will be retained even for newloc->inode, there by
preventing addition of file-size to parents in newpath.
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2689 (Quota: xattrs getting corrupted)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2689
|
|
|
|
|
|
|
|
|
|
| |
it takes up the new configuration values
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2777 (geo-replication config setting should have effect on already running sessions)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2777
|
|
|
|
|
|
|
|
| |
Signed-off-by: Pranith Kumar K <pranithk@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2794 (Alignment is not proper in profile volume info)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2794
|
|
|
|
|
|
|
|
|
|
| |
output of volume start/stop
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 1570 (geosync related changes)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1570
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Earlier rmdir would succeed on all up subvols, but fuse would get an
error if one of the subvol was down. In follow up lookup, self heal
would be triggered, and since st_mode would be 0, the permissions
would be bad. The behaviour now is to fail rmdir if subvol is down
Signed-off-by: shishir gowda <shishirng@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2591 (Directories changing to d--------- permission after trying to delete)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2591
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* provide a unified output in case of failure of geo-replication session
* provide a better formatting of the output geo-replication status command
* update the help option of geo-replication command
* fix few typos
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2536 (gsync service introspection)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2536
|
|
|
|
|
|
|
|
|
|
| |
to glusterd.
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2676 ([glusterfs-3.2.0qa7]: glusterfs server crashed)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2676
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
master is referred as a volume name rather than a URL scheme.
old syntax:
> volume geo-replication start :vol-foo /bar/boo
new syntax:
> volume geo-replication start vol-foo /bar/boo
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2786 (Having to prepend geo-replication master vol with colon spoils the UI)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2786
|
|
|
|
|
|
|
|
|
|
|
| |
the intelligence of determining the type of status command is now at glusterd,
rather then explicitly setting it while parsing at cli.
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2536 (gsync service introspection)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2536
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2736 (gsyncd hangs if crash occurs in the non-main thread)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2736
|
|
|
|
|
|
|
|
|
|
| |
graph validation
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2559 (provide two options in CLI for gluster volume gsync indexing <volname> <enable|disable>)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2559
|
|
|
|
|
|
|
|
| |
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2780 (geo-replication operations take too much time to complete)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2780
|
|
|
|
|
|
|
|
|
|
| |
in quota limit list.
Signed-off-by: Amar Tumballi <amar@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2732 (Quota: the local mount on server fails while using quota list command)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2732
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2649 (Quota: size field is not getting updated properly)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2649
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2784 ([glusterfs-3.2.0qa13]: memleak in marker)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2784
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2649 (Quota: size field is not getting updated properly)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2649
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra G <raghavendra@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2664 (Quota: recreating the volume on same bricks shows similar info as for earlier one using "list")
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2664
|
|
|
|
|
|
|
|
|
|
|
| |
This is useful for iterated polling of gsyncd's status, without having to
incoke gsyncd to find out about the pidfile.
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2780 (geo-replication operations take too much time to complete)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2780
|
|
|
|
|
|
|
|
|
|
|
| |
This reduces startup time for invocations other than master / slave role
(kind of which now proliferates glusterd)
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2780 (geo-replication operations take too much time to complete)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2780
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2772 ([glusterfs-3.2.0qa12]: excessive quota local NULL errors)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2772
|
|
|
|
|
|
|
|
| |
Signed-off-by: Raghavendra Bhat <raghavendrabhat@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2772 ([glusterfs-3.2.0qa12]: excessive quota local NULL errors)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2772
|
|
|
|
|
|
|
|
| |
Signed-off-by: Pranith Kumar K <pranithk@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2783 (Fix percentage latency and precisions of profile info)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2783
|
|
|
|
|
|
|
|
| |
Signed-off-by: Pranith Kumar K <pranithk@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2781 (geo-replication start error message)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2781
|
|
|
|
|
|
|
|
|
|
| |
so fsync call can succeed
Signed-off-by: Amar Tumballi <amar@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2772 ([glusterfs-3.2.0qa12]: excessive quota local NULL errors)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2772
|
|
|
|
|
|
|
|
|
|
|
|
| |
The final cleanup sequence + call to _exit, which was just done in the main
thread, now is called for in each thread when the thread crashes. Seems we
aren't left there hanging this way.
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2736 (gsyncd hangs if crash occurs in the non-main thread)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2736
|
|
|
|
|
|
|
|
|
|
|
|
| |
name, just the value
This intends to ease scripting (you can just pass on the value to xargs etc.)
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 1570 (geosync related changes)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=1570
|
|
|
|
|
|
|
|
|
|
| |
gsync session is active.
Signed-off-by: Kaushik BV <kaushikbv@gluster.com>
Signed-off-by: Anand Avati <avati@gluster.com>
BUG: 2559 (provide two options in CLI for gluster volume gsync indexing <volname> <enable|disable>)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2559
|