Perl 6 - the future is here, just unevenly distributed

IRC log for #gluster, 2016-07-14

| Channels | #gluster index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:01 caitnop joined #gluster
00:30 daMaestro joined #gluster
00:34 zat joined #gluster
00:35 zat has anyone tried to or could it not be a bad idea to “gluster” a system's system files between many computers?
00:35 zat the idea is to get something like a single OS shared between many computers
00:36 zat so an user would get the same home dir when logging at any workstation
00:36 zat same software, and so on...
00:43 pranithk1 joined #gluster
00:53 bsdjosh joined #gluster
00:53 shdeng joined #gluster
01:01 JoeJulian zat: Wouldn't be much different from nfs root, which has been done many times over. I think I've even seen an old blog post for a gluster root.
01:03 zat JoeJulian: so it's a proven concept?
01:06 squizzi_ joined #gluster
01:07 JoeJulian yep
01:07 JoeJulian Though ram's so cheap, I prefer just pxe booting and downloading a root image to a ram drive.
01:08 pranithk1 JoeJulian: Remember long time back you raised a bug not to let people upgrade when there are pending heals?
01:08 zat well, for my job, all these computers are provided by the education ministry so I can't really open them
01:09 zat I think I can make a massive RAID-10 like distributed FS???
01:10 pranithk1 JoeJulian: I am not sure how to get that upgrade script going... do you know? I think "gluster volume heal <volname> info" does print the files that need very accurately now.. so we can use this command to get it done. Wondering if it is possible for you to write that script. I am not good at writing these scripts :-), so wondering if you can help out.
01:12 JoeJulian I could, but what would be even smarter, imho, is a "gluster volume heal all pending" that returns a simple answer and an exit code.
01:13 JoeJulian Or even a "gluster server shutdown" that only does so when it's safe, ie no pending heals and server quorum will still be met.
01:15 JoeJulian zat: Assuming you mean replica 2 distributed, yes.
01:16 JoeJulian pranithk1: ping - just in case you timed out waiting two minutes for my reply. :)
01:16 zat I haven't used Gluster in years but I think you mean distribute in the network while ensuring the data is replicated twice?
01:16 zat well... once, for that matter.
01:16 pranithk1 JoeJulian: no I am here.. wait reading
01:17 JoeJulian Gluster volumes use replication and distribution, not raid. I try to avoid confusing the terms because they provide completely different performance.
01:17 pranithk1 JoeJulian: you mean a command which says if all the volumes in the cluster don't heal?
01:17 zat sure, that's why I said “RAID like”
01:17 zat lol
01:18 pranithk1 JoeJulian: Don't *need* heal
01:18 JoeJulian pranithk1:  "gluster volume heal all pending" = "No" exit code 0, "Yes" exit code 1
01:18 JoeJulian Or something like that.
01:19 pranithk1 JoeJulian: Hmm...
01:19 pranithk1 JoeJulian: got it
01:19 JoeJulian But I like my second idea better because it ensures quorum as well.
01:20 pranithk1 JoeJulian: Hmm... I think someone else asked for 1st command. May be that will get priority... I think server quorum should be deprecated.
01:20 pranithk1 JoeJulian: I don't see how it helps now that client-quorum is reasonably solid.
01:20 JoeJulian I see advantages to each. They solve different environment problems.
01:21 pranithk1 JoeJulian: May be for maintaining configuration consistency in glusterd. May be something to think about after glusterd2.0
01:21 JoeJulian +1
01:22 JoeJulian But if you get the simple heal state output, I'd be happy to write the .spec script.
01:23 JoeJulian btw... don't forget XML output.
01:23 pranithk1 JoeJulian: hey, XML output is already present :-)
01:23 JoeJulian For the new command, I mean.
01:23 pranithk1 JoeJulian: ah! I got it
01:24 jbrooks joined #gluster
01:24 pranithk1 JoeJulian: You mean first command right? "But if you get the simple heal state output,"
01:24 plarsen joined #gluster
01:24 muneerse joined #gluster
01:24 JoeJulian right
01:25 JoeJulian I will use that xml output in the saltstack gluster state.
01:28 muneerse2 joined #gluster
01:28 shyam joined #gluster
01:32 muneerse joined #gluster
01:37 Lee1092 joined #gluster
01:39 B21956 joined #gluster
01:47 ilbot3 joined #gluster
01:47 Topic for #gluster is now Gluster Community - http://gluster.org | Documentation - https://gluster.readthedocs.io/en/latest/ | Patches - http://review.gluster.org/ | Developers go to #gluster-dev | Channel Logs - https://botbot.me/freenode/gluster/ & http://irclog.perlgeek.de/gluster/
01:56 nishanth joined #gluster
02:04 pranithk1 left #gluster
02:51 poornimag joined #gluster
03:20 magrawal joined #gluster
03:26 kramdoss_ joined #gluster
03:48 nbalacha joined #gluster
03:49 atinm joined #gluster
03:51 itisravi joined #gluster
03:57 julim joined #gluster
04:02 nehar joined #gluster
04:18 RameshN joined #gluster
04:31 Gnomethrower joined #gluster
04:43 Manikandan joined #gluster
04:50 ppai joined #gluster
04:52 atalur joined #gluster
05:02 rafi joined #gluster
05:02 sakshi joined #gluster
05:02 aravindavk joined #gluster
05:04 prasanth joined #gluster
05:06 skoduri joined #gluster
05:07 shubhendu joined #gluster
05:07 ndarshan joined #gluster
05:14 aspandey joined #gluster
05:15 kshlm joined #gluster
05:21 javi404 joined #gluster
05:21 Bhaskarakiran joined #gluster
05:37 Apeksha joined #gluster
05:38 jiffin joined #gluster
05:45 hgowtham joined #gluster
05:48 nbalacha joined #gluster
05:48 kramdoss_ joined #gluster
05:50 harish_ joined #gluster
05:51 Javezim Anyone running Gluster on 16.04 w/ ZFS?
05:51 Javezim Ubuntu*
05:52 amye joined #gluster
05:55 kukulogy joined #gluster
05:55 harish_ joined #gluster
05:56 kukulogy Hi, I tried to mount `mount -t glusterfs gluster01:/strr /mnt` but it gives me an error. `mount: gluster01:/strr: Operation not supported by device` not sure if this is a problem on FreeBSD itself or from Glusterfs
06:00 jiffin kukulogy:I have never tried freebsd by my own.But i will try help debug in linux way  u have installed client packages, right?
06:01 jiffin Javezim: I have never tried, but can u please mention your issue?
06:01 nage joined #gluster
06:02 msvbhat joined #gluster
06:03 ashiq joined #gluster
06:03 khurram joined #gluster
06:06 atalur joined #gluster
06:08 kukulogy jiffin: Are you talking about the Gluster Native Client? I'm new so please bear with me. :\ I only tried to mount it on the server itself since I don't have a "client" yet.
06:09 jiffin kukulogy: yes. Thats k
06:09 sanoj joined #gluster
06:09 jiffin kukulogy: what all gluster packages have u installed?
06:11 kukulogy jiffin: http://www.freshports.org/net/glusterfs/ I installed glusterfs afaik.
06:11 glusterbot Title: FreshPorts -- net/glusterfs (at www.freshports.org)
06:13 jiffin kukulogy: do have any way to check packages installed in machine?
06:14 jiffin *in your machine
06:14 kukulogy jiffin: let me check. brb
06:14 jiffin and if possible please search gluster in it
06:14 satya4ever joined #gluster
06:15 kukulogy I only have this. glusterfs-3.7.6                GlusterFS distributed file system
06:16 jtux joined #gluster
06:16 kukulogy jiffin
06:16 nishanth joined #gluster
06:17 karnan joined #gluster
06:20 Muthu joined #gluster
06:20 devyani7_ joined #gluster
06:21 jiffin for fuse mount u require glusterfs-fuse package I guess
06:21 jiffin kukulogy:
06:22 jiffin kukulogy: I mean, for gluster native client mount , need to install glusterfs-fuse package
06:25 kukulogy jiffin: hmm okay. Perhaps the FUSE support is what I'm missing. Thanks for the insight. I'll research further.
06:25 Gnomethrower joined #gluster
06:28 Manikandan joined #gluster
06:29 rwheeler joined #gluster
06:32 ppai joined #gluster
06:32 kdhananjay joined #gluster
06:35 skoduri joined #gluster
06:36 anoopcs kukulogy, Can you try using the following command for mounting the volume:
06:36 anoopcs glusterfs --log-file=<logfile_name> --log-level=INFO --volfile-id=<volume_name> --volfile-server=<server_ip> <mount_point>
06:40 kukulogy anoopcs: http://dpaste.com/2E3XKRY so I think I'm missing fuse right?
06:40 glusterbot Title: dpaste: 2E3XKRY (at dpaste.com)
06:40 anoopcs kukulogy, Yes..
06:40 kukulogy anoopcs: I'm sorry lol
06:41 rastar joined #gluster
06:43 anoopcs kukulogy, You have to install FUSE manually..btw Which version are you using?
06:43 anoopcs *FreeBSD version
06:44 karthik_ joined #gluster
06:44 anoopcs kukulogy, https://wiki.freebsd.org/WhatsNew/FreeBSD10 confirms that FUSE is already part of FreeBSD 10.
06:44 glusterbot Title: WhatsNew/FreeBSD10 - FreeBSD Wiki (at wiki.freebsd.org)
06:44 kukulogy anoopcs: 10.2-RELEASE
06:45 anoopcs Hm.
06:50 anoopcs kukulogy, Can you try loading the FUSE kernel module?
07:03 * anoopcs could see that kldload command can be used to load kernel modules on FreeBSD.
07:03 kukulogy anoopcs: did I load it right? http://dpaste.com/35BYE62
07:03 glusterbot Title: dpaste: 35BYE62 (at dpaste.com)
07:04 anoopcs kkeithley, I don't an experience with FreeBSD. I would say `kldload fuse.ko` to be precise.
07:04 anoopcs Oops...wrong ping
07:04 glusterbot anoopcs: Please don't naked ping. http://blogs.gnome.org/markmc/2014/02/20/naked-pings/
07:05 kotreshhr joined #gluster
07:05 anoopcs kukulogy, Are you seeing the same error while mounting gluster volume after loading the module?
07:07 kukulogy anoopcs: I think it mounted successfully, I can see it now. /dev/fuse      126G    6.2G    110G     5%    /mnt/sdrr
07:07 kukulogy I mounted it using the command you gave me
07:09 anoopcs kukulogy, Verify that you can create files/directories via mount point and check in bricks.
07:10 pur_ joined #gluster
07:10 ndevos kukulogy: "mount -t glusterfs ..." (or whetever the FreeBSD equivalent is) would use /sbin/mount.glusterfs, maybe that is missing or fails in some other way
07:10 keiviw joined #gluster
07:12 kukulogy anoopcs: I can create file. I checked the file. it exists on Brick 1-3 but not on 4. I'm using Striped-Replicate. Is this normal?
07:15 kukulogy anoopcs: nvm, I found the file. Thank you!
07:15 hchiramm joined #gluster
07:18 anoopcs kukulogy, ndevos got it right here. You may be missing /sbin/mount.glusterfs which is a wrapper script for fuse mounting glusterfs volumes.
07:22 ivan_rossi joined #gluster
07:24 sbulage joined #gluster
07:29 nbalacha joined #gluster
07:29 kramdoss_ joined #gluster
07:30 arcolife joined #gluster
07:32 hchiramm joined #gluster
07:36 Saravanakmr joined #gluster
07:39 [Enrico] joined #gluster
07:42 fsimonce joined #gluster
07:48 Manikandan joined #gluster
08:04 karthik_ joined #gluster
08:04 ahino joined #gluster
08:07 derjohn_mobi joined #gluster
08:12 hackman joined #gluster
08:15 deniszh joined #gluster
08:16 ramky joined #gluster
08:31 kshlm joined #gluster
08:37 Javezim Hey All, having an issue with Gluster Samba VFS
08:37 Javezim http://paste.ubuntu.com/19345295/
08:37 glusterbot Title: Ubuntu Pastebin (at paste.ubuntu.com)
08:37 Javezim Appears there is no glusterfs.so file in this directory - /usr/lib/x86_64-linux-gnu/samba/vfs/glusterfs.so
08:37 Javezim So I cannot access the samba share from another machine
08:40 anoopcs Javezim, You need vfs module for glusterfs installed for this to work.
08:41 Javezim Where does one find that?
08:41 Javezim for 3.8
08:41 anoopcs Javezim, Hm.. I am not sure about 3.8..let me check.
08:42 anil joined #gluster
08:43 anoopcs Javezim, Yes..its there: https://launchpad.net/~monotek/+archive/ubuntu/samba-vfs-glusterfs-3.8
08:43 glusterbot Title: samba-vfs-glusterfs-3.8 : André Bauer (at launchpad.net)
08:43 Javezim Thanks
08:44 anoopcs Javezim, I think that PPA will provide glusterfs packages too. Make sure that you don't conflict with already installed packages.
08:44 anoopcs Javezim, Oh..No...nvm
08:45 Javezim Awesome, can access my shares now, Thanks @Anoopcs!
08:45 anoopcs Javezim, np
08:50 nehar joined #gluster
08:54 itisravi joined #gluster
09:02 prasanth joined #gluster
09:03 karthik_ joined #gluster
09:03 Bhaskarakiran_ joined #gluster
09:04 rafi1 joined #gluster
09:05 msvbhat joined #gluster
09:07 kshlm joined #gluster
09:07 Muthu joined #gluster
09:08 Seth_Karlo joined #gluster
09:16 ashiq joined #gluster
09:16 Manikandan joined #gluster
09:17 nehar joined #gluster
09:17 kotreshhr joined #gluster
09:19 aspandey joined #gluster
09:19 Saravanakmr joined #gluster
09:19 sanoj joined #gluster
09:24 msvbhat joined #gluster
09:29 kramdoss_ joined #gluster
09:47 rafi joined #gluster
09:48 itisravi_ joined #gluster
09:51 robb_nl joined #gluster
10:00 ashiq joined #gluster
10:01 Saravanakmr joined #gluster
10:01 sanoj joined #gluster
10:01 jwd joined #gluster
10:01 kotreshhr joined #gluster
10:03 karthik_ joined #gluster
10:04 prasanth joined #gluster
10:20 bfoster joined #gluster
10:25 atalur joined #gluster
10:35 harish_ joined #gluster
10:35 ira joined #gluster
10:37 johnmilton joined #gluster
10:38 muneerse2 joined #gluster
10:39 sakshi joined #gluster
10:43 muneerse joined #gluster
10:45 Klas what version of liburcu is required for 3.7.13 on the client side?
10:46 muneerse2 joined #gluster
10:49 nbalacha joined #gluster
10:50 nehar joined #gluster
10:50 atinm joined #gluster
10:50 muneerse joined #gluster
10:51 anoopcs Klas, >=0.8 I guess.
10:51 Klas hmm, damn it
10:52 Klas means I need to add a newer version to our trusty repos
10:52 Klas thanks =)
10:52 anoopcs Klas, Wait...Let me verify it once more.
10:52 Klas sure thing
10:52 kotreshhr joined #gluster
10:53 Klas it sounds reasonable, cause that is the one parth which seems to differ noticeable between my failing trusty and my xenial
10:53 Klas (xenial servers, xenial and trusty clients)
10:55 Klas hmm, actually, no, I just made a mistake, it seems to work fine
10:55 anoopcs Klas, Yeah.. I think 0.7 will also work.
10:55 Klas oh, nice
10:56 Klas what does liburcu do, meaning, how can I verify?
10:56 anoopcs Klas, Verify what?
10:56 Klas meaning how can I verify that it actually works as intended ;)?
10:56 Klas I do not know for what function liburcu is needed
10:58 anoopcs Klas, glusterd uses it. See http://liburcu.org/
10:58 glusterbot Title: Userspace RCU (at liburcu.org)
10:59 Klas yup, I've read that, but I'm uncertain which part of glusterd is using it
10:59 Klas what will break if I don't have it and what will break with incorrect version, so to speak?
11:00 Klas mounting and writing works fine without it, that's why I'm asking ;)
11:00 anoopcs Klas, Which version are you using?
11:00 anoopcs liburcu version?
11:01 Klas default trusty, 0.7.12-0ubuntu2, default xenial, 0.9.1-3, and no version at all are the test scenarios =)
11:02 Klas btw, I've got nothing against rtfm, as long as I'm reading a manual that will answer the question =)
11:05 Klas from random googling it seems to deal with consistency of data when multiple clients are writing, so it seems important
11:05 khurram joined #gluster
11:07 khurram hi there; i have a glusterfs setup on pis; i am able to mount volume as long as client is not behind NAT; i need to mount it on Docker Containers (for devs) and those containers gets setup behind NAT; can i do something on server that they are able to mount gluster?
11:08 khurram i have rpc-auth-allow-insecure on in /etc/glusterfs/glusterd.vol and server.allow-insecure on for said volume; but still they are unable to mount
11:08 Klas anoopcs: the documentation primarily lists it as a build requirement, not a client requirement, but I can't find client requirements in general otoh
11:10 anoopcs Klas, We check for urcu/cds.h at the least
11:10 anoopcs If we can't find a version og liburcu-cds >=0.7
11:11 Klas on client as well?
11:11 Klas cause mounts seem to be working fine without it
11:11 Klas and writes
11:12 anoopcs AFAIK, This is specific to glusterd and not for clients.
11:13 anoopcs Klas, kshlm can provide more input on liburcu.
11:13 Klas anoopcs: ah, great, I was just basing the requirements on the requirements in ubuntus packages
11:14 Klas thanks for the answers to my uninformed questions =)
11:14 itisravi joined #gluster
11:14 Klas kshlm: can you verify that liburcu is unneccessary dependancy for a glusterfs client?
11:15 nehar joined #gluster
11:17 kshlm Klas, it is not needed for any client side binaries. Only glusterd uses liburcu.
11:17 Klas sweet, thanks =)
11:17 Klas is there any good place to look for client-side dependancies?
11:24 Manikandan joined #gluster
11:24 guhcampos joined #gluster
11:25 guhcampos_ joined #gluster
11:27 khurram hi there; i have a glusterfs setup on pis; i am able to mount volume as long as client is not behind NAT; i need to mount it on Docker Containers (for devs) and those containers gets setup behind NAT; can i do something on server that they are able to mount gluster?
11:27 khurram i have rpc-auth-allow-insecure on in /etc/glusterfs/glusterd.vol and server.allow-insecure on for said volume; but still they are unable to mount!
11:27 khurram log from client: 0-gv-client-0: SETVOLUME on remote-host failed: Authentication failed, sending AUTH_FAILED event
11:27 anoopcs Klas, I found this: apt-cache rdepends glusterfs-client (I am not an expert on apt package management)
11:29 Klas anoopcs: ah, yes, I didn't mean ubuntus view on things, I just want to verify if they where wrong with anything else except liburcu4 which they listed as a dep
11:29 Klas or, oh yeah
11:29 Klas you mean the .deb you guys are supplying
11:29 kshlm khurram, Have you set any "auth.allow' or "auth.reject" options on the volume?
11:30 anoopcs Klas, I was talking from packaging perspective and I see that you were referring to Ubuntu.
11:30 anoopcs Thus the `apt` command.
11:30 kshlm The error you see indicates that the servers are rejecting connections beacause auth failed.
11:30 Klas anoopcs: yeah, I meant in general since I'm rolling my own ;)
11:30 skoduri joined #gluster
11:31 khurram no; no auth settings on volume
11:31 anoopcs Klas, Different dependencies are defined for various sub-packages while building for each distro.
11:32 kshlm Could you share any of the brick logs?
11:32 aspandey joined #gluster
11:32 kshlm Clients behind nat shold work as long as auth isn't being used.
11:32 Klas anoopcs: yup, that part I do understand, I'm interested in looking at the necessary libs and seeing if ubuntu is giving me "more crap I don't need" when using their packages =)
11:33 kshlm auth breaks down with nat as the servers only get the nat gateway address, and not the client address for incoming connections.
11:33 khurram brick log: [addr.c:143:gf_auth] 0-auth/addr: client is bound to port 53011 which is not privileged
11:33 khurram [authenticate.c:239:gf_authenticate] 0-auth: no authentication module is interested in accepting remote-client (null)
11:34 kshlm khurram,  Have you set `rpc-auth-allow-insecure on` on the volume?
11:35 kshlm BTW, which version of gluster are you using?
11:35 khurram yes; rpc-auth-allow-insecure is on and server.allow-insecure is also on
11:35 kshlm If it's greater than 3.7.3, gluster should allow insecure ports by default.
11:36 khurram unfortunately its 3.5.2; gluster is running on raspberry pis
11:36 darshan joined #gluster
11:36 kshlm That is a really old version! And currently unsupported.
11:36 kshlm Where are you getting the packages from?
11:37 kshlm Aside from that, have you restarted the volume?
11:38 khurram i used apt-get on raspbian (raspberry pi)
11:38 khurram i think i restarted the volume; i try that again
11:39 ndarshan joined #gluster
11:40 redhat joined #gluster
11:40 kotreshhr joined #gluster
11:41 khurram thanks kshlm; its working now!
11:41 khurram may be i forgot restarting volume, my bad :|
11:41 kshlm Cool!
11:41 khurram how can i get latest binaries for raspbian (ARM) ?
11:42 kshlm I don't think we're building arm packages yet.
11:42 khurram yes; i tried find on download.gluster.org; but its not there
11:42 kshlm If there is Fedora or CentOS images for Pi, they may have it.
11:43 khurram ok, i will try looking around
11:43 kkeithley You can't get the latest, but at least there's 3.6.5 for Raspian Jessie at http://download.gluster.org/pub/gluster/glusterfs/3.6/3.6.5/Raspbian/
11:43 glusterbot Title: Index of /pub/gluster/glusterfs/3.6/3.6.5/Raspbian (at download.gluster.org)
11:43 pdrakeweb joined #gluster
11:44 kshlm kkeithley, Thanks.
11:44 khurram thanks kkeithley!
11:44 kshlm Any reason why we stopped at that point?
11:44 kkeithley 3.7.3 at http://download.gluster.org/pub/gluster/glusterfs/3.7/3.7.3/Raspbian/jessie/
11:44 glusterbot Title: Index of /pub/gluster/glusterfs/3.7/3.7.3/Raspbian/jessie (at download.gluster.org)
11:45 jwd joined #gluster
11:45 kkeithley or http://download.gluster.org/pub/gluster/glusterfs/3.7/3.7.4/Raspbian/
11:45 glusterbot Title: Index of /pub/gluster/glusterfs/3.7/3.7.4/Raspbian (at download.gluster.org)
11:46 kshlm BTW, feodra builds glusterfs for armv7hl packages of gluster.
11:46 kkeithley I only have so much copious spare time
11:47 kshlm kkeithley, Thanks for putting them out.
11:47 kshlm kkeithley++
11:47 glusterbot kshlm: kkeithley's karma is now 27
11:47 ndarshan joined #gluster
11:47 kkeithley Fedora builds don't work on RaspPi1 or RaspPi2.  If you have RaspPi3s then you can run Fedora and, indeed, gluster packages are available in Fedora.
11:48 kshlm Is that because of the arm32 and arm64 differences?
11:49 kkeithley no.  RaspPi1 and RaspPi2 are armv6, Fedora only supports armv7.  RaspPi3 finally switched to armv7.
11:50 kkeithley For old RaspPi there was a Fedora remix called Pidora.
11:50 kshlm I need to get a few of these single board computers to play around with.
11:51 anmol joined #gluster
11:51 kshlm Just checked the ubuntu ppa, they have arm packages as well.
11:56 kkeithley which PPA?
12:01 kkeithley I was only aware of i686 and amd64 pkgs in the PPA.  Is Launchpad building .debs for secondary arches?
12:06 atinm joined #gluster
12:08 kshlm kkeithley, That was my mistake.
12:08 kshlm I was browsing the repo on the web.
12:09 kshlm http://ppa.launchpad.net/gluster/glusterfs-3.8/ubuntu/dists/xenial/main/ , this link lists a lot of archs.
12:09 glusterbot Title: Index of /gluster/glusterfs-3.8/ubuntu/dists/xenial/main (at ppa.launchpad.net)
12:09 kshlm But only i686 and amd64 packages are present in the pool.
12:10 natarej joined #gluster
12:20 Wizek joined #gluster
12:26 prasanth joined #gluster
12:33 nehar joined #gluster
12:45 anmol joined #gluster
13:06 julim joined #gluster
13:11 rastar joined #gluster
13:15 msvbhat joined #gluster
13:22 rastar joined #gluster
13:30 Manikandan joined #gluster
13:30 Manikandan left #gluster
13:31 jiffin1 joined #gluster
13:33 kramdoss_ joined #gluster
13:37 unclemarc joined #gluster
13:39 pdrakewe_ joined #gluster
13:43 dnunez joined #gluster
13:45 jkroon joined #gluster
13:51 nishanth joined #gluster
13:52 jiffin1 joined #gluster
13:53 nbalacha joined #gluster
13:53 skylar joined #gluster
13:53 squizzi joined #gluster
13:55 jiffin joined #gluster
14:01 arcolife joined #gluster
14:09 dnunez joined #gluster
14:14 BitByteNybble110 joined #gluster
14:15 Jacob843 joined #gluster
14:17 msvbhat joined #gluster
14:19 Jacob843 joined #gluster
14:25 pdrakeweb joined #gluster
14:26 plarsen joined #gluster
14:32 Seth_Karlo joined #gluster
14:36 gluster-newbie joined #gluster
14:38 gluster-newbie So we are planning to change our "performance.io-thread-count" from 16 to 64.  Do we need to change the other "performance.<priority>-prio-threads" properties, where priority = [high,normal,low,least]?  What do these properties do?
14:46 farhorizon joined #gluster
14:47 cloph unfortunately documentation on the performance tuning parameters is very rare, most googling only finds some get variables dump without any reasoning as to why values are chosen.. (so sorry, only user myself, so cannot really help on this one)
15:00 rastar joined #gluster
15:09 wushudoin joined #gluster
15:17 kotreshhr joined #gluster
15:20 plarsen joined #gluster
15:26 jwd joined #gluster
15:29 guhcampos joined #gluster
15:47 kotreshhr joined #gluster
15:48 kotreshhr left #gluster
15:58 kdhananjay joined #gluster
16:03 karnan joined #gluster
16:21 nehar joined #gluster
16:25 kdhananjay joined #gluster
16:38 ivan_rossi left #gluster
16:56 farhorizon joined #gluster
16:58 ashiq joined #gluster
17:02 anmol joined #gluster
17:05 squizzi_ joined #gluster
17:09 shyam joined #gluster
17:17 ashiq_ joined #gluster
17:20 unclemarc joined #gluster
17:21 ben453 joined #gluster
17:25 [o__o] joined #gluster
17:30 jiffin joined #gluster
17:43 farhorizon joined #gluster
17:45 hchiramm joined #gluster
17:45 farhorizon joined #gluster
17:47 guhcampos_ joined #gluster
17:51 farhorizon joined #gluster
17:55 farhoriz_ joined #gluster
17:58 rafi joined #gluster
17:59 shyam joined #gluster
17:59 julim joined #gluster
18:02 msvbhat joined #gluster
18:04 shaunm joined #gluster
18:08 ahino joined #gluster
18:09 kpease joined #gluster
18:10 alvinstarr joined #gluster
18:14 farhoriz_ Getting an interesting report from split brain (gluster volume heal storage info split-brain) - it identifies a single file as split brained, but only on one of two bricks & that file no longer exists on either volume.  Not sure why this would come up in split brain
18:14 ira joined #gluster
18:14 nishanth joined #gluster
18:15 rastar joined #gluster
18:16 farhoriz_ Since the file didn't exist anywhere, I tried deleting the metadata for it & running full heal, but the split brain report still claims it is an issue
18:17 rastar joined #gluster
18:28 [o__o] joined #gluster
18:40 Siavash joined #gluster
18:50 cliluw joined #gluster
19:06 deniszh joined #gluster
19:26 derjohn_mobi joined #gluster
19:35 [o__o] joined #gluster
19:42 Siavash joined #gluster
19:42 Siavash joined #gluster
19:49 julim joined #gluster
20:17 guhcampos joined #gluster
20:18 plarsen joined #gluster
20:36 pocketprotector joined #gluster
20:39 hchiramm joined #gluster
20:42 v12aml joined #gluster
20:58 DV joined #gluster
21:05 bluenemo joined #gluster
21:10 hchiramm joined #gluster
21:29 kotreshhr joined #gluster
21:34 hchiramm joined #gluster
21:35 amye joined #gluster
21:41 MattGanderup joined #gluster
21:47 ira joined #gluster
22:06 hchiramm joined #gluster
22:33 hchiramm joined #gluster
22:49 shyam joined #gluster
23:58 hackman joined #gluster

| Channels | #gluster index | Today | | Search | Google Search | Plain-Text | summary