| Time |
Nick |
Message |
| 00:01 |
|
cristov joined #gluster |
| 00:43 |
|
sjm joined #gluster |
| 00:44 |
|
vpshastry joined #gluster |
| 00:48 |
|
verdurin joined #gluster |
| 00:54 |
|
gmcwhistler joined #gluster |
| 01:11 |
|
sauce joined #gluster |
| 01:22 |
|
harish joined #gluster |
| 01:40 |
|
vpshastry joined #gluster |
| 01:51 |
|
sjm left #gluster |
| 01:51 |
|
_2_panda3 joined #gluster |
| 01:51 |
_2_panda3 |
hi |
| 01:51 |
glusterbot |
_2_panda3: Despite the fact that friendly greetings are nice, please ask your question. Carefully identify your problem in such a way that when a volunteer has a few minutes, they can offer you a potential solution. These are volunteers, so be patient. Answers may come in a few minutes, or may take hours. If you're still in the channel, someone will eventually offer an answer. |
| 01:52 |
_2_panda3 |
:) |
| 01:52 |
|
gildub joined #gluster |
| 02:11 |
|
haomaiwa_ joined #gluster |
| 02:12 |
|
bala joined #gluster |
| 02:21 |
|
harish joined #gluster |
| 02:26 |
|
Pupeno_ joined #gluster |
| 02:26 |
|
haomaiw__ joined #gluster |
| 02:35 |
Peter3 |
JoeJulian: I am seeing the df vs du issue again and this time i see the du -b is diff between nfs and glusterfs mount |
| 02:37 |
Peter3 |
http://pastie.org/9370048 |
| 02:37 |
glusterbot |
Title: #9370048 - Pastie (at pastie.org) |
| 02:46 |
|
ThatGraemeGuy joined #gluster |
| 02:54 |
|
bharata-rao joined #gluster |
| 02:59 |
|
aravindavk joined #gluster |
| 03:36 |
|
kumar joined #gluster |
| 03:37 |
|
atinmu joined #gluster |
| 03:39 |
|
nbalachandran joined #gluster |
| 03:52 |
|
RameshN joined #gluster |
| 03:52 |
|
aravindavk joined #gluster |
| 03:55 |
|
itisravi joined #gluster |
| 04:01 |
|
plarsen joined #gluster |
| 04:06 |
|
_ndevos joined #gluster |
| 04:15 |
|
kanagaraj joined #gluster |
| 04:18 |
|
vpshastry joined #gluster |
| 04:20 |
|
shubhendu joined #gluster |
| 04:26 |
|
deepakcs joined #gluster |
| 04:31 |
|
nishanth joined #gluster |
| 04:37 |
|
dtrainor joined #gluster |
| 04:46 |
|
sahina joined #gluster |
| 04:49 |
|
nshaikh joined #gluster |
| 04:52 |
|
ndarshan joined #gluster |
| 05:01 |
|
sputnik1_ joined #gluster |
| 05:06 |
|
vpshastry joined #gluster |
| 05:18 |
|
ppai joined #gluster |
| 05:18 |
|
prasanth joined #gluster |
| 05:26 |
|
kdhananjay joined #gluster |
| 05:32 |
|
gmcwhist_ joined #gluster |
| 05:49 |
|
kshlm joined #gluster |
| 05:49 |
|
hagarth joined #gluster |
| 05:50 |
|
saurabh joined #gluster |
| 05:54 |
|
davinder17 joined #gluster |
| 05:54 |
|
stickyboy joined #gluster |
| 05:54 |
|
vkoppad joined #gluster |
| 06:31 |
|
ricky-ti1 joined #gluster |
| 06:32 |
|
vimal joined #gluster |
| 06:32 |
|
karnan joined #gluster |
| 06:37 |
|
spandit joined #gluster |
| 06:38 |
|
gmcwhist_ joined #gluster |
| 06:43 |
|
vu joined #gluster |
| 06:45 |
|
monotek joined #gluster |
| 07:00 |
|
davinder16 joined #gluster |
| 07:03 |
|
ctria joined #gluster |
| 07:04 |
|
rastar joined #gluster |
| 07:05 |
|
rastar_ joined #gluster |
| 07:05 |
|
eseyman joined #gluster |
| 07:08 |
|
keytab joined #gluster |
| 07:14 |
|
ekuric joined #gluster |
| 07:14 |
|
RaSTar joined #gluster |
| 07:15 |
glusterbot |
New news from newglusterbugs: [Bug 1114604] [FEAT] Improve SSL support <https://bugzilla.redhat.com/show_bug.cgi?id=1114604> |
| 07:15 |
|
RaSTar joined #gluster |
| 07:21 |
|
nbalachandran joined #gluster |
| 07:26 |
|
ppai joined #gluster |
| 07:29 |
|
pvh_sa joined #gluster |
| 07:33 |
|
fsimonce joined #gluster |
| 07:36 |
|
qdk joined #gluster |
| 07:45 |
glusterbot |
New news from newglusterbugs: [Bug 1117655] 0-mem-pool: invalid argument with fio --thread <https://bugzilla.redhat.com/show_bug.cgi?id=1117655> |
| 07:58 |
|
rjoseph joined #gluster |
| 08:00 |
|
aravindavk joined #gluster |
| 08:02 |
|
fdsfsdf joined #gluster |
| 08:02 |
|
mshadle joined #gluster |
| 08:13 |
|
hybrid512 joined #gluster |
| 08:29 |
|
calum_ joined #gluster |
| 08:34 |
|
lalatenduM joined #gluster |
| 08:48 |
|
hagarth joined #gluster |
| 09:04 |
|
ppai joined #gluster |
| 09:12 |
|
meghanam joined #gluster |
| 09:19 |
|
RameshN joined #gluster |
| 09:25 |
|
andreask joined #gluster |
| 09:27 |
|
meghanam_ joined #gluster |
| 09:27 |
|
sahina joined #gluster |
| 09:28 |
|
shubhendu joined #gluster |
| 09:31 |
|
ninkotech joined #gluster |
| 09:32 |
|
ninkotech__ joined #gluster |
| 09:35 |
|
bala1 joined #gluster |
| 09:39 |
|
kdhananjay joined #gluster |
| 09:43 |
|
nishanth joined #gluster |
| 09:48 |
|
haomaiwang joined #gluster |
| 09:52 |
|
ndarshan joined #gluster |
| 09:53 |
|
meghanam_ joined #gluster |
| 09:53 |
|
meghanam joined #gluster |
| 10:01 |
|
haomaiwa_ joined #gluster |
| 10:04 |
|
kshlm joined #gluster |
| 10:09 |
|
ninkotech joined #gluster |
| 10:09 |
|
deepakcs joined #gluster |
| 10:09 |
|
ninkotech__ joined #gluster |
| 10:16 |
|
kshlm joined #gluster |
| 10:27 |
|
Slashman joined #gluster |
| 10:40 |
|
nbalachandran joined #gluster |
| 10:45 |
glusterbot |
New news from newglusterbugs: [Bug 1077406] Striped volume does not work with VMware esxi v4.1, 5.1 or 5.5 <https://bugzilla.redhat.com/show_bug.cgi?id=1077406> |
| 10:47 |
|
gildub joined #gluster |
| 10:56 |
|
edward1 joined #gluster |
| 11:01 |
|
meghanam joined #gluster |
| 11:01 |
|
meghanam_ joined #gluster |
| 11:01 |
|
karnan joined #gluster |
| 11:07 |
pvh_sa |
hey there, if I'm doing multiple bricks on a single server, should they be in different lvols and different filesystems? or can I just create them as different directories of the same filesystem? |
| 11:14 |
|
bene2 joined #gluster |
| 11:14 |
|
ppai joined #gluster |
| 11:14 |
|
pdrakeweb joined #gluster |
| 11:15 |
pvh_sa |
seems if you use the same FS then free space doesn't show up right, so I'm making 1 lvol per brick |
| 11:19 |
|
sahina joined #gluster |
| 11:21 |
|
calum_ joined #gluster |
| 11:27 |
|
ndarshan joined #gluster |
| 11:30 |
_ndevos |
pvh_sa: some use the 'one filesystem for many bricks' approach and call it 'thin provisioning' |
| 11:32 |
ndevos |
pvh_sa: but yes, the size of the volume depends on the size of the underlying filesystem that is used for the bricks, so if you want 'correct' or 'understandable' values, use one filesystem per brick |
| 11:33 |
|
nishanth joined #gluster |
| 11:36 |
|
shubhendu joined #gluster |
| 11:37 |
|
LebedevRI joined #gluster |
| 11:38 |
|
bala1 joined #gluster |
| 11:39 |
pvh_sa |
ndevos, thanks. |
| 11:39 |
pvh_sa |
so on my bricks, do I run glusterd *and* glusterfsd ? what's the difference between them? |
| 11:40 |
|
qdk joined #gluster |
| 11:40 |
ndevos |
glusterd is the management daemon, it handles all the requests from the 'gluster' commandline, and passes the volume layout to clients that mount/use the volume |
| 11:40 |
ndevos |
glusterd also starts additional gluster services, like glusterfsd, the gluster-nfs server, etc |
| 11:41 |
ndevos |
glusterfsd is a process that listens on a tcp-port, and handles all I/O for a specific brick |
| 11:41 |
|
harish_ joined #gluster |
| 11:41 |
ndevos |
glusterfs processes are clients (fuse-mount, nfs-server, self-heal, ...) and 1st connect to glusterd, and then the client know to which bricks to talk |
| 11:43 |
pvh_sa |
ok thanks, so... both need to be running then? because gluster will talk to glusterd to do 'create' and then clients will talk to glusterfsd |
| 11:43 |
ndevos |
yes, when you create+start a volume, glusterd starts glusterfsd |
| 11:44 |
ndevos |
when you mount a volume, glusterfs-fuse connects to glusterd, gets the volume layout, and uses that volume layout to connect to the bricks (glusterfsd) |
| 11:44 |
pvh_sa |
ok, so there is no need to enable glusterfsd manually (as in chkconfig) because glusterd will start that? |
| 11:44 |
ndevos |
yes, thats correct |
| 11:45 |
ndevos |
the glusterfsd sevice or systemd job is used to stop the glusterfsd processes on reboot/shutdown/etc |
| 11:45 |
ndevos |
s/sevice/service/ |
| 11:45 |
glusterbot |
What ndevos meant to say was: the glusterfsd service or systemd job is used to stop the glusterfsd processes on reboot/shutdown/etc |
| 12:00 |
|
kanagaraj joined #gluster |
| 12:01 |
|
RameshN joined #gluster |
| 12:07 |
|
theron joined #gluster |
| 12:08 |
|
theron joined #gluster |
| 12:09 |
|
pdrakeweb joined #gluster |
| 12:12 |
|
itisravi joined #gluster |
| 12:13 |
|
itisravi joined #gluster |
| 12:27 |
|
tdasilva joined #gluster |
| 12:28 |
|
sputnik1_ joined #gluster |
| 12:31 |
|
RameshN_ joined #gluster |
| 12:34 |
|
pdrakeweb joined #gluster |
| 12:42 |
|
Thilam joined #gluster |
| 12:44 |
|
chirino joined #gluster |
| 12:45 |
|
stickyboy joined #gluster |
| 12:48 |
|
julim joined #gluster |
| 13:00 |
|
japuzzo joined #gluster |
| 13:04 |
|
_Bryan_ joined #gluster |
| 13:06 |
|
TvL2386 joined #gluster |
| 13:14 |
|
sahina joined #gluster |
| 13:16 |
|
ndarshan joined #gluster |
| 13:16 |
glusterbot |
New news from newglusterbugs: [Bug 1117822] Tracker bug for GlusterFS 3.6.0 <https://bugzilla.redhat.com/show_bug.cgi?id=1117822> |
| 13:16 |
|
hagarth joined #gluster |
| 13:17 |
kkeithley |
,,(ports} |
| 13:17 |
glusterbot |
I do not know about 'ports}', but I do know about these similar topics: 'ports' |
| 13:17 |
kkeithley |
,,(ports) |
| 13:17 |
glusterbot |
glusterd's management port is 24007/tcp and 24008/tcp if you use rdma. Bricks (glusterfsd) use 24009 & up for <3.4 and 49152 & up for 3.4. (Deleted volumes do not reset this counter.) Additionally it will listen on 38465-38467/tcp for nfs, also 38468 for NLM since 3.3.0. NFS also depends on rpcbind/portmap on port 111 and 2049 since 3.4. |
| 13:19 |
|
deeville joined #gluster |
| 13:21 |
|
shubhendu joined #gluster |
| 13:22 |
|
hagarth left #gluster |
| 13:22 |
|
hagarth joined #gluster |
| 13:22 |
|
nbalachandran joined #gluster |
| 13:22 |
Thilam |
hi there, I've just update my 3.5 to 3.5.1 glusterfs and I always have an issue when I whant to enable quota |
| 13:23 |
Thilam |
gluster volume quota projets enable |
| 13:23 |
deeville |
Is there a way to specify hosts to root-squash? Or is there no way around this global setting? |
| 13:23 |
Thilam |
quota: Could not start quota auxiliary mount |
| 13:23 |
Thilam |
Quota command failed. Please check the cli logs for more details |
| 13:23 |
pvh_sa |
hey there, how should I specify my mount command so that it is doesn't have a single point of failure (the server named in the mount command)? |
| 13:23 |
Thilam |
in logs [2014-07-09 13:20:09.066609] E [cli-cmd-volume.c:1036:gf_cli_create_auxiliary_mount] 0-cli: Failed to create auxiliary mount directory /var/run/gluster/projets/. Reason : No such file or directory |
| 13:23 |
|
sjm joined #gluster |
| 13:24 |
Thilam |
when i create the folder it's ok, but I suppose it's not a normal behaviour |
| 13:24 |
Thilam |
in fact It's the /var/run/gluster folder which is missing |
| 13:25 |
|
nishanth joined #gluster |
| 14:12 |
|
ilbot3 joined #gluster |
| 14:12 |
|
Topic for #gluster is now Gluster Community - http://gluster.org | Patches - http://review.gluster.org/ | Developers go to #gluster-dev | Channel Logs - https://botbot.me/freenode/gluster/ & http://irclog.perlgeek.de/gluster/ |
| 14:13 |
|
bennyturns joined #gluster |
| 14:15 |
ndevos |
Thilam: maybe you can figure out what <gfid:7d74ef56-b073-421e-b9e6-1c1d1927cfd0> corresponds too, and heal it |
| 14:16 |
|
pvh_sa joined #gluster |
| 14:16 |
glusterbot |
New news from newglusterbugs: [Bug 1117851] DHT :- data loss - file is missing on renaming same file from multiple client at same time <https://bugzilla.redhat.com/show_bug.cgi?id=1117851> |
| 14:16 |
ndevos |
Thilam: <gfid:7d74ef56-b073-421e-b9e6-1c1d1927cfd0> should be a symlink on the brick, on an other brick, check <brick>/.glusterfs/7d/74/7d74ef56-b073-421e-b9e6-1c1d1927cfd0 |
| 14:17 |
|
chirino_m joined #gluster |
| 14:17 |
ndevos |
Thilam: the target of the symlink, is the name of the directory on the volume, you should be able to heal that, by 'stat <mountpoint>/target/of/symlink' |
| 14:18 |
ndevos |
the <mountpoint> should be a fuse-mount, not the mountpoint of the brick |
| 14:18 |
|
mortuar joined #gluster |
| 14:19 |
|
davinder16 joined #gluster |
| 14:19 |
ndevos |
Thilam: dependin on the directory-depth of the symlink, you may need to heal the parent dirs first, like 'stat <mountpoint>/target', 'stat <mountpoint>/target/of' |
| 14:20 |
|
cjanbanan1 joined #gluster |
| 14:20 |
|
VerboEse joined #gluster |
| 14:21 |
|
jobewan joined #gluster |
| 14:23 |
cjanbanan1 |
Where can I find information how GlusterFS reads from a replicated brick? According to the profiling data, it seems to choose only one of bricks, but how is this decision taken and does it last until that particular brick is removed? |
| 14:23 |
|
theron joined #gluster |
| 14:25 |
|
rwheeler joined #gluster |
| 14:27 |
|
ricky-ticky1 joined #gluster |
| 14:27 |
|
diegows joined #gluster |
| 14:35 |
Thilam |
thx ndevos |
| 14:36 |
Thilam |
but do you have seen my previous message about quota activation ? |
| 14:36 |
Thilam |
15:22 |
| 14:37 |
Thilam |
I tought this pb have to be solved in 3.5.1 release |
| 14:37 |
Thilam |
thought sorry |
| 14:38 |
ndevos |
Thilam: right, enabling quota should create that directory, you can file a bug for that and add glusterfs-3.5.2 in the blocks field (need to enable 'advanced' mode for reporting) |
| 14:38 |
glusterbot |
https://bugzilla.redhat.com/enter_bug.cgi?product=GlusterFS |
| 14:39 |
Thilam |
k thx |
| 14:40 |
Thilam |
on my way |
| 14:49 |
Thilam |
sorry, I've enable the "advanced" report but I don't find the field to specify glusterfs-3.5.2 |
| 14:51 |
semiosis |
Thilam: 3.5.2 is not out yet. use pre-release |
| 14:57 |
kkeithley |
3.5.2 is a choice for the Version: |
| 14:58 |
Thilam |
yeah but it's not in the advanced fields |
| 14:58 |
Thilam |
like ndevos asked me, I was searching for a target release field |
| 14:58 |
Thilam |
nevermind |
| 14:59 |
kkeithley |
after you create the BZ you can update and put 3.5.2 in the Target Release |
| 14:59 |
semiosis |
ah so it is |
| 15:00 |
ndevos |
Thilam: I meant the text-field called 'blocks', you can add "glusterfs-3.5.2" in there |
| 15:00 |
JustinClift |
*** Gluster Community Meeting is NOW. #gluster-meeting on irc.freenode.net *** |
| 15:01 |
Thilam |
there is no text field called blocks :) |
| 15:01 |
Thilam |
even if I am in advanced view |
| 15:02 |
sauce |
my two glusterfs servers in AWS crashed at the same time :( this is the second time this has happened with VERY light usage. i can't even SSH in |
| 15:02 |
sauce |
AWS monitoring shows both had CPU go to 100% about 30 minutes ago |
| 15:03 |
|
jdarcy joined #gluster |
| 15:03 |
|
lmickh joined #gluster |
| 15:05 |
|
andreask joined #gluster |
| 15:07 |
|
anoopcs joined #gluster |
| 15:09 |
ndevos |
Thilam: I have that field on the bottom of my enter_bug.cgi page... |
| 15:09 |
ndevos |
Thilam: you can also try to set it by using https://bugzilla.redhat.com/enter_bug.cgi?product=GlusterFS&blocked=glusterfs-3.5.2 |
| 15:09 |
glusterbot |
Title: Log in to Red Hat Bugzilla (at bugzilla.redhat.com) |
| 15:15 |
|
anoopcs left #gluster |
| 15:16 |
Thilam |
I'll use your 2nd link, on the bottom of my enter_bug.cgi, the last text field is Description :) |
| 15:16 |
Thilam |
ho no, the last is Bug-ID |
| 15:17 |
|
vpshastry joined #gluster |
| 15:17 |
|
LebedevRI joined #gluster |
| 15:18 |
|
gmcwhistler joined #gluster |
| 15:19 |
Thilam |
hey, go the blocks field one I've submitted my bug :) |
| 15:20 |
Thilam |
-one +once |
| 15:22 |
|
P0w3r3d joined #gluster |
| 15:23 |
|
bala joined #gluster |
| 15:27 |
|
cjanbanan1 left #gluster |
| 15:27 |
|
cjanbanan joined #gluster |
| 15:28 |
|
marbu joined #gluster |
| 15:32 |
|
bala1 joined #gluster |
| 15:36 |
|
bala joined #gluster |
| 15:40 |
|
Peter3 joined #gluster |
| 15:44 |
|
LebedevRI joined #gluster |
| 15:46 |
|
Eco_ joined #gluster |
| 15:47 |
glusterbot |
New news from newglusterbugs: [Bug 1117888] Problem when enabling quota : Could not start quota auxiliary mount <https://bugzilla.redhat.com/show_bug.cgi?id=1117888> || [Bug 1117886] Gluster not resolving hosts with IPv6 only lookups <https://bugzilla.redhat.com/show_bug.cgi?id=1117886> |
| 15:54 |
|
ron-slc joined #gluster |
| 15:57 |
|
deeville joined #gluster |
| 15:57 |
ndevos |
Thilam: nice, you seem to have added the blocker after filing the bug? glad to know that works at least :) |
| 16:01 |
Thilam |
yes |
| 16:03 |
Thilam |
strange, but it's ok now :) |
| 16:05 |
|
doo joined #gluster |
| 16:06 |
|
jcsp joined #gluster |
| 16:12 |
|
MacWinner joined #gluster |
| 16:17 |
glusterbot |
New news from newglusterbugs: [Bug 1117921] Wrong releaseversion picked up when doing 'make -C extras/LinuxRPM glusterrpms' <https://bugzilla.redhat.com/show_bug.cgi?id=1117921> |
| 16:18 |
|
tg2 joined #gluster |
| 16:20 |
|
MacWinner joined #gluster |
| 16:21 |
|
MacWinner joined #gluster |
| 16:29 |
|
Mo__ joined #gluster |
| 16:31 |
|
davinder16 joined #gluster |
| 16:37 |
|
elico joined #gluster |
| 16:37 |
|
cjanbanan joined #gluster |
| 16:52 |
|
coredump joined #gluster |
| 17:01 |
|
pvh_sa joined #gluster |
| 17:13 |
|
kanagaraj joined #gluster |
| 17:14 |
|
vimal joined #gluster |
| 17:15 |
|
rturk joined #gluster |
| 17:17 |
glusterbot |
New news from newglusterbugs: [Bug 1117951] Use C-locale for numerics (strtod and friends) <https://bugzilla.redhat.com/show_bug.cgi?id=1117951> |
| 17:29 |
|
plarsen joined #gluster |
| 17:32 |
|
ricky-ti1 joined #gluster |
| 17:36 |
|
cristov joined #gluster |
| 17:37 |
|
qdk joined #gluster |
| 17:38 |
anotheral |
is it normal for heals to be very very slow? This one has only restored about 11GB of data after 15 hours |
| 17:39 |
|
bfoster joined #gluster |
| 17:42 |
jiffe98 |
so if I ls in a directory and get a whole bunch of IO errors and files with question marks, how does one go about fixing that? |
| 17:43 |
JoeJulian_ |
jiffe98: Start with "gluster volume status" |
| 17:44 |
jiffe98 |
JoeJulian_: everything looks good there, bricks and nfs servers all running |
| 17:44 |
JoeJulian_ |
anotheral: You can always just walk the directory tree from a client mount. It shouldn't be that slow, though, unless you've got very busy servers. |
| 17:44 |
JoeJulian_ |
jiffe98: check that directory on the bricks? |
| 17:46 |
anotheral |
JoeJulian_: ah, that will force the heal to catch up? |
| 17:47 |
JoeJulian_ |
anotheral: Anything that's needs healed /should/ be healed by the client that way. You can verify by looking in the client log. |
| 17:47 |
anotheral |
this is a distributed replicated 2x2 cluster, and one of the bricks up and died |
| 17:47 |
anotheral |
(four bricks, one per node) |
| 17:47 |
JoeJulian_ |
s/node/server |
| 17:48 |
anotheral |
ok |
| 17:48 |
anotheral |
ah, it's already at 20GB now! |
| 17:48 |
anotheral |
thanks, JoeJulian_ |
| 17:48 |
JoeJulian_ |
You're welcome |
| 17:49 |
anotheral |
25GB now - muuuuch better. |
| 17:51 |
jiffe98 |
files are there but the dates on them are different |
| 17:55 |
jiffe98 |
we rsync one of each mirror every night, sounds like they tried to restore someone by rsyncing in the other direction |
| 17:57 |
|
deeville joined #gluster |
| 18:04 |
jiffe98 |
apparently that process doesn't work, rsyncing both mirrors so they both match and I still get errors |
| 18:05 |
Peter3 |
what do we do with failures in volume rebalance? |
| 18:07 |
|
calum_ joined #gluster |
| 18:08 |
|
LebedevRI joined #gluster |
| 18:10 |
|
Eco_ joined #gluster |
| 18:11 |
|
hchiramm_ joined #gluster |
| 18:17 |
tg2 |
rm -rf * |
| 18:21 |
|
andreask joined #gluster |
| 18:28 |
purpleidea |
Peter3: ignore tg2 he's a bot |
| 18:31 |
|
plarsen joined #gluster |
| 18:31 |
Peter3 |
ok |
| 18:31 |
semiosis |
purpleidea: oh no |
| 18:31 |
Peter3 |
any idea what we should do with the failures on volume rebalance |
| 18:31 |
Peter3 |
? |
| 18:32 |
purpleidea |
semiosis: what's up? |
| 18:32 |
semiosis |
this is the second bot of yours i'm having to ignore ;) |
| 18:35 |
|
pureflex joined #gluster |
| 18:38 |
|
cjanbanan joined #gluster |
| 18:49 |
|
qdk joined #gluster |
| 18:53 |
purpleidea |
semiosis: lol no it's a real person that i know irl. but he's trolling! lol |
| 18:53 |
purpleidea |
semiosis: i think it's safe to keep him /ignore-d lol |
| 18:54 |
semiosis |
we'll see |
| 18:55 |
* JoeJulian |
can always kickban him... |
| 18:55 |
|
chirino joined #gluster |
| 18:55 |
Peter3 |
you guys talking about me? ;( |
| 18:55 |
JoeJulian |
Failures on a rebalance are usually: the target drive has less free space than the source drive |
| 18:56 |
purpleidea |
Peter3: no, you're good |
| 18:56 |
Peter3 |
:) |
| 18:56 |
Peter3 |
thanks! |
| 18:57 |
Peter3 |
JoeJulian: if the source drive has more space then target, then why would it try rebalance? |
| 18:57 |
purpleidea |
Peter3: random unrelated, believe it or not, but I know another "peter3" :P |
| 18:57 |
JoeJulian |
@lucky dht misses are expensive |
| 18:57 |
glusterbot |
JoeJulian: http://joejulian.name/blog/dht-misses-are-expensive/ |
| 18:57 |
JoeJulian |
Peter3: Read the first part of that to learn how dht works. ^ |
| 18:58 |
Peter3 |
thanks! |
| 19:03 |
|
rotbeard joined #gluster |
| 19:18 |
|
coredump joined #gluster |
| 19:19 |
|
ekuric joined #gluster |
| 19:21 |
|
cultavix joined #gluster |
| 19:22 |
|
cultavix joined #gluster |
| 19:22 |
cultavix |
hi all |
| 19:25 |
semiosis |
hi |
| 19:25 |
glusterbot |
semiosis: Despite the fact that friendly greetings are nice, please ask your question. Carefully identify your problem in such a way that when a volunteer has a few minutes, they can offer you a potential solution. These are volunteers, so be patient. Answers may come in a few minutes, or may take hours. If you're still in the channel, someone will eventually offer an answer. |
| 19:25 |
semiosis |
cultavix: ^^^ |
| 19:25 |
cultavix |
semiosis :) |
| 19:26 |
cultavix |
I'm very happy with my gluster cluster so far, just got it setup at home now |
| 19:26 |
cultavix |
now looking to expand it with no downtime |
| 19:26 |
cultavix |
and take advantage of some of its features |
| 19:26 |
cultavix |
like snapshotting |
| 19:26 |
cultavix |
geo-replication, etc |
| 19:27 |
cultavix |
does each node have to have the same size disk ? |
| 19:27 |
semiosis |
well, no, but that's usually a good idea |
| 19:27 |
cultavix |
Ive got 2x servers with 100GB drives each |
| 19:28 |
cultavix |
could I add for example, 2x servers with 500GB drives now |
| 19:28 |
cultavix |
and combine everything |
| 19:28 |
cultavix |
not really sure what the best way is |
| 19:28 |
semiosis |
glusterfs distributes *files* evenly among the bricks (not bytes) so you could run out of space on your smaller bricks before your larger ones |
| 19:28 |
cultavix |
ah ok, that makes sense |
| 19:28 |
semiosis |
you could split the 500GB drives into 5 x 100GB bricks (with LVM, or partitions) |
| 19:28 |
cultavix |
so I've sort of limited myself to 100GB right now |
| 19:29 |
cultavix |
ah ok, that sounds cool |
| 19:29 |
semiosis |
replicate between the servers & distribute over the bricks |
| 19:29 |
cultavix |
I could do like a 2 x 2 = 4 ? |
| 19:29 |
cultavix |
and just get the right sizes via lvm |
| 19:30 |
semiosis |
something like that... though if you have 5 x 100 on one pair of servers and 1 x 100 on another it would be 2 x 6 |
| 19:30 |
cultavix |
ah ok, so I would create a new brick for each 100GB partition/lvm |
| 19:30 |
semiosis |
yep |
| 19:30 |
cultavix |
and I can do that on the current 2x servers or shall I create a few more ? |
| 19:30 |
cultavix |
(this is all for just testing btw, so its just for me to learn) |
| 19:30 |
semiosis |
you can have many bricks per server |
| 19:31 |
cultavix |
I'd like use as much of the disk as I can, with at least 1 replica |
| 19:31 |
cultavix |
so if one fails, then it's no problem |
| 19:31 |
|
Paul-C joined #gluster |
| 19:33 |
Peter3 |
Why would this happen and do I need to worry about? |
| 19:33 |
Peter3 |
http://pastie.org/9372368 |
| 19:33 |
cultavix |
ok, here is what im going to do, im going to create 2x new servers with 2x 100gb disks, id like to expand my diskspace by 100gb |
| 19:33 |
glusterbot |
Title: #9372368 - Pastie (at pastie.org) |
| 19:34 |
semiosis |
cultavix: well, there are some benefits to expanding the bricks in-place, with lvm, rather than adding more bricks |
| 19:35 |
semiosis |
cultavix: if you add-brick then you need to rebalance, which is an expensive operation |
| 19:35 |
semiosis |
cultavix: if you just expand existing bricks you can avoid that |
| 19:35 |
cultavix |
but I'd like to add more complexity, at least 2 more servers |
| 19:36 |
semiosis |
cultavix: to add servers you can either add-brick (req's rebalance) or move existing bricks to the new server |
| 19:36 |
semiosis |
if you plan ahead with many bricks, that is |
| 19:36 |
cultavix |
ah ok, I can move them onto new servers... |
| 19:36 |
cultavix |
thats handy! |
| 19:36 |
semiosis |
the replace-brick command |
| 19:36 |
semiosis |
which only moves the data on that brick, not all data on all bricks like rebalance |
| 19:37 |
cultavix |
ok cool, thanks buddy... I'll get cracking.... I may come back later on tonight or tomorrow, you guys have been very helpful |
| 19:37 |
cultavix |
I've only got 1 brick |
| 19:37 |
semiosis |
good luck |
| 19:37 |
cultavix |
cheers :) |
| 19:37 |
* semiosis |
biased against rebalance |
| 19:38 |
|
cjanbanan joined #gluster |
| 19:38 |
Peter3 |
semiosis: huh? |
| 19:39 |
semiosis |
i dont trust it |
| 19:40 |
Peter3 |
you mean you do not trust rebalance? |
| 19:41 |
semiosis |
that is what i mean |
| 19:41 |
Peter3 |
i am still getting missing space between df and du |
| 19:41 |
Peter3 |
i wonder if rebalance would fix it |
| 19:41 |
semiosis |
i dont trust du either |
| 19:41 |
semiosis |
the list of things i do not trust is extensive |
| 19:43 |
|
tdasilva joined #gluster |
| 19:43 |
|
firemanxbr joined #gluster |
| 19:49 |
|
Peter3 joined #gluster |
| 19:53 |
|
KORG joined #gluster |
| 20:02 |
stickyboy |
semiosis: :P |
| 20:24 |
|
SpeeR joined #gluster |
| 20:29 |
|
coredump|br joined #gluster |
| 20:31 |
|
chirino joined #gluster |
| 20:37 |
|
cjanbanan joined #gluster |
| 20:43 |
|
coredump joined #gluster |
| 20:55 |
|
coredump|br joined #gluster |
| 21:10 |
|
ninkotech__ joined #gluster |
| 21:17 |
|
hchiramm_ joined #gluster |
| 21:21 |
|
pureflex joined #gluster |
| 21:34 |
|
cultavix joined #gluster |
| 21:42 |
|
dtrainor joined #gluster |
| 22:00 |
|
cjanbanan joined #gluster |
| 22:00 |
|
sjm left #gluster |
| 22:11 |
|
sonicrose joined #gluster |
| 22:12 |
sonicrose |
hi all. is this a good place to hire a gluster expert/freelancer to take a look over my gluster setup? |
| 22:13 |
JoeJulian |
JustinClift: ^ |
| 22:15 |
sonicrose |
i really need this storage to run at top notch, and being that I've set it all up by myself with nothing to go by, im 100% that if i can get a gluster expert(s) to review my setup we can make it 100% better than it is now |
| 22:15 |
sonicrose |
its worth a few hundred dollars to me |
| 22:15 |
|
klaas joined #gluster |
| 22:16 |
JoeJulian |
I would imagine a few thousand dollars at the least, but that's just my expectation. |
| 22:18 |
|
gmcwhistler joined #gluster |
| 22:19 |
|
coredump joined #gluster |
| 22:24 |
sonicrose |
thousands huh? dang |
| 22:38 |
|
cjanbanan joined #gluster |
| 22:51 |
|
cultavix joined #gluster |
| 22:53 |
|
chirino joined #gluster |
| 22:53 |
|
fidevo joined #gluster |
| 22:58 |
Eco_ |
if anyone wants to take a gander, just rolled out the new gluster site at www.gluster.org |
| 22:59 |
JoeJulian |
prdy |
| 22:59 |
Eco_ |
the old site still exists so if it makes folks sad we can always roll back ;) |
| 22:59 |
JoeJulian |
Can I try to break it?!?! |
| 22:59 |
Eco_ |
please do |
| 22:59 |
Eco_ |
will write a proper blog post about the changes but wanted to get it out first |
| 23:00 |
JoeJulian |
I'll spin up a few thousand VM's and put it through some real testing! |
| 23:00 |
JoeJulian |
... actually, I won't. I have to figure this other thing out first. |
| 23:03 |
Eco_ |
JoeJulian, lol |
| 23:04 |
Eco_ |
its middleman based now so it should at least take longer to fall over than the previous site |
| 23:22 |
|
pureflex joined #gluster |
| 23:34 |
|
necrogami joined #gluster |
| 23:35 |
|
gildub joined #gluster |
| 23:44 |
|
gmcwhistler joined #gluster |
