#ceph IRC Log

Index

IRC Log for 2013-09-04

Timestamps are in GMT/BST.

[0:01] * marrusl (~mark@209-150-43-182.c3-0.wsd-ubr2.qens-wsd.ny.cable.rcn.com) has joined #ceph
[0:02] <mikedawson> joshd: I'm just trying to effectively document the problem, not sure what else I can do to help get this fixed (other than dusting off my feeble C++ skills)
[0:04] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) Quit (Remote host closed the connection)
[0:10] * sleinen1 (~Adium@2001:620:0:25:7c2c:7154:96aa:4930) Quit (Quit: Leaving.)
[0:10] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) has joined #ceph
[0:15] * sjm (~sjm@38.122.20.226) has joined #ceph
[0:15] * MarkN1 (~nathan@197.204.233.220.static.exetel.com.au) has joined #ceph
[0:18] * MarkN (~nathan@142.208.70.115.static.exetel.com.au) has joined #ceph
[0:18] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) Quit (Quit: Leaving.)
[0:19] <sagewk> gregaf: sjust: https://github.com/ceph/ceph/pull/563
[0:20] * markbby (~Adium@168.94.245.1) Quit (Quit: Leaving.)
[0:21] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) Quit (Ping timeout: 480 seconds)
[0:23] * MarkN1 (~nathan@197.204.233.220.static.exetel.com.au) Quit (Ping timeout: 480 seconds)
[0:28] <sjust> sagewk: what locks does the objecter hold when it calls into the client callbacks?
[0:29] <sagewk> none; we use a Finisher
[0:29] <sjust> k
[0:30] * AfC (~andrew@2407:7800:200:1011:2ad2:44ff:fe08:a4c) has joined #ceph
[0:37] <sjust> sagewk: left a minor comment on the librados part, and some questions about the OpContext vs Objecter callback lifetimes
[0:38] <sjust> sagewk: also, I stupidly commented on the commits rather than on the overall diff
[0:38] <sagewk> np
[0:39] * MarkN (~nathan@142.208.70.115.static.exetel.com.au) has left #ceph
[0:40] * dmsimard1 (~Adium@ap02.wireless.co.mtl.iweb.com) has joined #ceph
[0:42] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) Quit (Remote host closed the connection)
[0:42] * dmsimard (~Adium@108.163.152.2) Quit (Read error: Operation timed out)
[0:46] <Kioob> Hi
[0:47] <Kioob> little mistake on http://ceph.com/docs/master/dev/generatedocs/ : the git clone domain is "ceph.com", not "ceph" : git clone git://ceph.com/ceph.git
[0:49] * dmsimard1 (~Adium@ap02.wireless.co.mtl.iweb.com) Quit (Ping timeout: 480 seconds)
[0:52] * sprachgenerator (~sprachgen@c-50-141-192-36.hsd1.il.comcast.net) has joined #ceph
[0:53] * jluis (~joao@89.181.152.211) has joined #ceph
[0:53] * ChanServ sets mode +o jluis
[0:54] <sagewk> sjust: repushed wip-copyfrom
[0:55] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[0:55] <dmick> Kioob: actually it should probably be github.com/ceph/ceph. Will fix.
[0:56] <sjust> sagewk: looks good to me
[0:56] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) Quit (Quit: Ja odoh a vi sta 'ocete...)
[0:56] <Kioob> oh, I always use ceph.com for git clone
[0:57] <Kioob> well, you're right again : https://github.com/ceph/ceph.git
[0:58] * PerlStalker (~PerlStalk@2620:d3:8000:192::70) Quit (Quit: ...)
[0:59] * joao (~joao@89.181.152.211) Quit (Ping timeout: 480 seconds)
[1:01] <dmick> Kioob: refresh that page :0
[1:07] <Kioob> thanks
[1:08] <Kioob> note : I was searching for the ceph-client.git, which I found at ceph.com/ceph-client.git
[1:10] <dmick> also github/ceph/ceph-client
[1:10] <Kioob> ok, thanks
[1:11] * The_Bishop (~bishop@2001:470:50b6:0:742a:9fef:cf4e:6181) has joined #ceph
[1:17] * sjm (~sjm@38.122.20.226) has joined #ceph
[1:24] * carif (~mcarifio@pool-96-233-32-122.bstnma.fios.verizon.net) Quit (Ping timeout: 480 seconds)
[1:26] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Remote host closed the connection)
[1:33] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) has joined #ceph
[1:34] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Ping timeout: 480 seconds)
[1:39] * alphe (~alphe@0001ac6f.user.oftc.net) has joined #ceph
[1:40] <alphe> hello every one !
[1:40] * gaveen (~gaveen@175.157.69.118) Quit (Ping timeout: 480 seconds)
[1:40] <alphe> I have several questions to make around ceph data storage and S3
[1:41] * mtanski (~mtanski@69.193.178.202) Quit (Ping timeout: 480 seconds)
[1:42] * ross__ (~ross@60.208.111.209) has joined #ceph
[1:43] <alphe> I have several questions to make around ceph object gateway, ceph block device and S3
[1:43] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) has joined #ceph
[1:44] <alphe> ok so I dis a Ceph Object Gateway following the documentation and filling the blanks and I got an s3 like storage working
[1:45] <alphe> I could even connect to it windows clients but still I have some questions that needs to be answered
[1:45] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[1:45] <alphe> for example in Ceph Block device what is the "page" amount that is given to a pool ...
[1:46] <alphe> for example radosgw need to have a .rgw pool so you create it using osd create .rgw then should follow the number of page
[1:47] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[1:47] <alphe> when I do ceph -s I see in the stat returned that there is an amount of pages display are those the same pages that osd pools will use ?
[1:47] <alphe> what means giving 1 page to the .users pool ?
[1:47] <alphe> does that means i will only be able to store 1 user ?
[1:48] <alphe> (guess not since logically 1 page is a large amount of GB)
[1:48] <alphe> in the documentation around the pools needed to be created for a s3 like data store
[1:49] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[1:49] <alphe> you only have a list of pools to create and no infomations about the amount of pages you have to attribute ...
[1:50] * nwat (~nwat@eduroam-237-79.ucsc.edu) Quit (Ping timeout: 480 seconds)
[1:50] <alphe> the SSL layer for the httpd /fastcgi/radosgw needed by s3 amazon protocole was half explain in the documentation (wiki)
[1:50] <alphe> and for the pools needed it is the same
[1:51] <alphe> you have the list of the name of pools to create you have nothing conserning their weight or size
[1:52] <alphe> think I should have that talk in mail ...
[1:54] <alphe> can the data stored through a s3 like gateway can be seen by a samba ?
[1:55] <alphe> (I think they dont since they use completly different mechanism ...)
[1:55] * AfC (~andrew@2407:7800:200:1011:2ad2:44ff:fe08:a4c) Quit (Quit: Leaving.)
[1:55] * AfC1 (~andrew@2407:7800:200:1011:2ad2:44ff:fe08:a4c) has joined #ceph
[1:55] * sprachgenerator (~sprachgen@c-50-141-192-36.hsd1.il.comcast.net) Quit (Quit: sprachgenerator)
[1:58] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Ping timeout: 480 seconds)
[1:58] * BillK (~BillK-OFT@203-59-161-9.dyn.iinet.net.au) Quit (Ping timeout: 480 seconds)
[2:00] * sjm (~sjm@38.122.20.226) has joined #ceph
[2:02] <MACscr> alphe: lol, come one. using the mailing list for something like that
[2:04] * mschiff_ (~mschiff@port-46124.pppoe.wtnet.de) has joined #ceph
[2:04] * BillK (~BillK-OFT@58-7-131-166.dyn.iinet.net.au) has joined #ceph
[2:05] <alphe> Macsrc ?
[2:05] <alphe> hum it is important ...
[2:06] <alphe> to me sure
[2:06] <alphe> not for the whole planet since I don t find answers in the documentations I need to ask to people that will be able to answer me
[2:06] * clayb (~kvirc@proxy-nj2.bloomberg.com) Quit (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/)
[2:07] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[2:07] <rturk> alphe: people are working on samba plugins, but they use the block/file interfaces to my knowledge, not the gateway
[2:07] <alphe> rturk yes ... I know I deployed both
[2:08] <alphe> I use samba over ceph-fuse or over cephfs
[2:08] <alphe> but samba is damb slow ...
[2:09] <alphe> it is pratical from you windows client you can map the samba share as a network drive super easyly but that is convenient for a bunch of secretaries that want to have a comon data store with their boss ...
[2:10] <alphe> I need to store replications of hard drive recoveries ... wich is a totally different world ...
[2:10] <alphe> and speed means monney in that world ...
[2:10] <alphe> we tryed previously samba over zfs it was darn slow too ...
[2:11] <alphe> and not really tolerent to faillures
[2:11] <alphe> we turst ceph is the right way since 1 year now. And understand me in a linux monolitic world it is fantastic
[2:11] * mschiff (~mschiff@port-54664.pppoe.wtnet.de) Quit (Ping timeout: 480 seconds)
[2:12] <alphe> in a mixed technologies evironement it is a pain samba offers me a 250kb data transfer when it comes to transfer tons of small files
[2:12] * tnt (~tnt@91.176.35.47) Quit (Ping timeout: 480 seconds)
[2:13] <alphe> in a linux pure environement the same takes means a 40MB data rate ...
[2:14] <alphe> so yes it is difficult for me to show my boss that we are on the right track ...
[2:15] <alphe> to make better the data transfere rate I opted for a s3 like rados gateway and with the s3 clients I tested I could reach 8MB for that same task which isn t as bad as samba
[2:15] * xarses (~andreww@204.11.231.50.static.etheric.net) Quit (Remote host closed the connection)
[2:16] <alphe> sure all this would be better if there was a windows native client to talk directly to ceph cluster ...
[2:17] <alphe> but i will not write it ... and I assume really few people could do such a task ...
[2:18] <alphe> so radosgw s3 like is odd ... I was really enthousiast about it until the last update show me that after updating ceph and regionmap I couldn t retreive the data ..
[2:19] * sagelap (~sage@2607:f298:a:607:ea03:9aff:febc:4c23) Quit (Ping timeout: 480 seconds)
[2:21] <alphe> since I have a limited documentation a limited understanding of what i am doing with radosgw I couldn t test the swift/openstack windows clients
[2:22] <alphe> maybe those clients would have shown a 16MB or 30MB continous upload rate to the ceph cluster ...
[2:23] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[2:24] <alphe> most of the windows openstack clients are the same as the s3 amazon clients so I am sceptical ...
[2:28] * sjm (~sjm@38.122.20.226) has joined #ceph
[2:30] * sagelap (~sage@2600:1012:b028:5ac3:a1d8:6962:87a3:483b) has joined #ceph
[2:30] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[2:34] * yy-nm (~Thunderbi@125.122.52.139) has joined #ceph
[2:36] * yanzheng (~zhyan@jfdmzpr01-ext.jf.intel.com) has joined #ceph
[2:41] * sjm (~sjm@38.122.20.226) has joined #ceph
[2:47] * rturk is now known as rturk-away
[2:48] * alram (~alram@38.122.20.226) Quit (Quit: leaving)
[2:53] * nerdtron (~kenneth@202.60.8.252) has joined #ceph
[2:59] * bandrus (~Adium@cpe-76-95-217-129.socal.res.rr.com) Quit (Quit: Leaving.)
[3:00] <nerdtron> hi all is it safe to update ceph on a running cluster?
[3:00] <nerdtron> i want to go from 0.61.3 to 0.61.8
[3:00] * freedomhui (~freedomhu@106.3.103.180) has joined #ceph
[3:01] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Remote host closed the connection)
[3:02] <dpippenger> nerdtron: in my experience it's been a safe procedure unless the release notes indicate a specific incompatibility with the upgrade path.
[3:03] <nerdtron> dpippenger, thanks!, just ceph-deploy install node1 node2 node3 right??
[3:03] <dpippenger> nerdtron: the only one I'm aware of at the moment is there *was* an issue with upgradeing from 0.56.x to 0.67.x, but it was resolved in 0.67.2
[3:04] <dpippenger> I haven't used the ceph-deploy install process myself, maybe someone else here has more experience with that. I use saltstack to upgrade the packages and restart osd instances in an orderly fashion.
[3:04] <sglwlb> hi, all, in list bucket type, if i remove the last item in list, just cause the movement of the last item?
[3:04] * sagelap (~sage@2600:1012:b028:5ac3:a1d8:6962:87a3:483b) Quit (Read error: Connection reset by peer)
[3:05] <sglwlb> data movement
[3:08] <alphe> 0.61-3 to 0.61-8 is ok
[3:08] <dpippenger> nerdtron: it doesn't appear that the ceph-deploy install actually restarts any of the processes, so after you have executed it you will still have to coordinate a restart of processes to get yourself on to the new code. http://ceph.com/docs/next/install/upgrading-ceph/
[3:08] <alphe> dpippenger yes
[3:08] <nerdtron> thanks, i'll take a look at it
[3:09] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Ping timeout: 480 seconds)
[3:09] * smiley_ (~smiley@pool-173-73-0-53.washdc.fios.verizon.net) has joined #ceph
[3:17] * angdraug (~angdraug@204.11.231.50.static.etheric.net) Quit (Quit: Leaving)
[3:19] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) has joined #ceph
[3:21] <nerdtron> thanks dpippenger! successfully upgraded the ceph mons and osd
[3:24] * Cube (~Cube@12.248.40.138) Quit (Quit: Leaving.)
[3:27] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[3:28] * freedomhui (~freedomhu@106.3.103.180) Quit (Quit: Leaving...)
[3:28] * grepory (~Adium@50-115-70-146.static-ip.telepacific.net) Quit (Quit: Leaving.)
[3:32] * freedomhui (~freedomhu@106.3.103.140) has joined #ceph
[3:34] <alphe> ok leaving goodbye
[3:34] * alphe (~alphe@0001ac6f.user.oftc.net) Quit (Quit: Leaving)
[3:34] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[3:36] * xarses (~andreww@c-71-202-167-197.hsd1.ca.comcast.net) has joined #ceph
[3:36] * kyann (~oftc-webi@AMontsouris-652-1-224-208.w92-128.abo.wanadoo.fr) Quit (Remote host closed the connection)
[3:38] * diegows (~diegows@190.190.11.42) Quit (Ping timeout: 480 seconds)
[3:43] * sjm (~sjm@38.122.20.226) has joined #ceph
[3:46] * KevinPerks (~Adium@cpe-066-026-239-136.triad.res.rr.com) Quit (Quit: Leaving.)
[3:48] * dpippenger (~riven@tenant.pas.idealab.com) Quit (Ping timeout: 480 seconds)
[3:57] * freedomhui (~freedomhu@106.3.103.140) Quit (Quit: Leaving...)
[4:00] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) Quit (Quit: Leaving.)
[4:01] * freedomhui (~freedomhu@211.155.113.212) has joined #ceph
[4:09] * imjustmatthew (~imjustmat@pool-173-53-100-217.rcmdva.fios.verizon.net) Quit (Remote host closed the connection)
[4:09] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[4:10] * alram (~alram@cpe-76-167-50-51.socal.res.rr.com) has joined #ceph
[4:15] * yy-nm (~Thunderbi@125.122.52.139) Quit (Remote host closed the connection)
[4:16] * yy-nm (~Thunderbi@125.122.52.139) has joined #ceph
[4:20] * Cube (~Cube@66-87-65-158.pools.spcsdns.net) has joined #ceph
[4:23] * KevinPerks (~Adium@cpe-066-026-239-136.triad.res.rr.com) has joined #ceph
[4:37] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) Quit (Remote host closed the connection)
[4:47] * shang (~ShangWu@175.41.48.77) has joined #ceph
[4:53] * cofol1986 (~xwrj@110.90.119.113) Quit (Read error: Connection reset by peer)
[4:53] * freedomhui (~freedomhu@211.155.113.212) Quit (Read error: Connection reset by peer)
[4:53] * freedomhui (~freedomhu@117.79.232.248) has joined #ceph
[5:03] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) Quit (Quit: Leaving.)
[5:04] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) has joined #ceph
[5:04] * freedomhui (~freedomhu@117.79.232.248) Quit (Ping timeout: 480 seconds)
[5:06] * fireD_ (~fireD@93-139-182-124.adsl.net.t-com.hr) has joined #ceph
[5:07] * fireD (~fireD@93-139-160-160.adsl.net.t-com.hr) Quit (Ping timeout: 480 seconds)
[5:09] * sprachgenerator (~sprachgen@c-50-141-192-36.hsd1.il.comcast.net) has joined #ceph
[5:34] * freedomhui (~freedomhu@106.3.103.180) has joined #ceph
[5:34] * Cube (~Cube@66-87-65-158.pools.spcsdns.net) Quit (Quit: Leaving.)
[5:46] * sprachgenerator (~sprachgen@c-50-141-192-36.hsd1.il.comcast.net) Quit (Quit: sprachgenerator)
[5:58] * KevinPerks (~Adium@cpe-066-026-239-136.triad.res.rr.com) Quit (Quit: Leaving.)
[6:01] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[6:11] * yy-nm (~Thunderbi@125.122.52.139) Quit (Quit: yy-nm)
[6:15] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[6:20] * haomaiwang (~haomaiwan@112.193.130.196) has joined #ceph
[6:23] * alram (~alram@cpe-76-167-50-51.socal.res.rr.com) Quit (Ping timeout: 480 seconds)
[6:28] * KevinPerks (~Adium@cpe-066-026-239-136.triad.res.rr.com) has joined #ceph
[6:32] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[6:34] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) Quit (Quit: Leaving.)
[6:37] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) has joined #ceph
[6:39] * KevinPerks (~Adium@cpe-066-026-239-136.triad.res.rr.com) Quit (Ping timeout: 480 seconds)
[6:39] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) has joined #ceph
[6:40] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Read error: Operation timed out)
[6:40] * sleinen1 (~Adium@2001:620:0:25:b521:770a:f969:a10d) has joined #ceph
[6:47] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) Quit (Ping timeout: 480 seconds)
[6:58] * Cube (~Cube@66-87-65-158.pools.spcsdns.net) has joined #ceph
[7:00] * jlogan2 (~Thunderbi@2600:c00:3010:1:1::40) has joined #ceph
[7:02] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[7:05] <sage> yanzheng: ping!
[7:05] <yanzheng> hi
[7:06] <yanzheng> sage, ?
[7:06] <sage> hey- do you have time to look at majianpeng's latest 2 patches?
[7:06] * jlogan (~Thunderbi@2600:c00:3010:1:1::40) Quit (Ping timeout: 480 seconds)
[7:07] <yanzheng> will do this afternoon
[7:07] <sage> also, i'm putting together an initial pull request that will include the tested stuff and some rbd fixes so that they can go to stable kernels quickly (missed 3.11). the fscache and d_prune stuff is waiting on david howells and al, tho
[7:07] <sage> is there anythign you have that isn't in the testing branch yet?
[7:07] <sage> thanks!
[7:08] <yanzheng> nothing else
[7:13] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) Quit (Ping timeout: 480 seconds)
[7:16] * The_Bishop (~bishop@2001:470:50b6:0:742a:9fef:cf4e:6181) Quit (Quit: Wer zum Teufel ist dieser Peer? Wenn ich den erwische dann werde ich ihm mal die Verbindung resetten!)
[7:17] * sleinen1 (~Adium@2001:620:0:25:b521:770a:f969:a10d) Quit (Quit: Leaving.)
[7:17] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) has joined #ceph
[7:21] * sjusthm (~sam@24-205-35-233.dhcp.gldl.ca.charter.com) has joined #ceph
[7:25] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) Quit (Ping timeout: 480 seconds)
[7:25] * Cube (~Cube@66-87-65-158.pools.spcsdns.net) Quit (Ping timeout: 480 seconds)
[7:30] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[7:33] * topro (~topro@host-62-245-142-50.customer.m-online.net) Quit (Quit: Konversation terminated!)
[7:43] * huangjun (~kvirc@111.172.32.75) has joined #ceph
[7:44] <huangjun> does the total number of files in /dev/disk/by-path change?
[7:48] * julian (~julianwa@125.70.133.115) has joined #ceph
[7:50] <huangjun> when i used the ceph-deploy on my centos 6.4 and there are always 3 files in /dev/disk/by-path
[7:53] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[7:56] * yy-nm (~Thunderbi@125.122.52.139) has joined #ceph
[7:58] <nerdtron> currently the journal of each osd (3 osd) is on the drive itself. Will i get faster performance if I put convert one hard disk as journal and the other two will remain as osd?
[8:05] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[8:18] * Cube (~Cube@cpe-76-95-217-129.socal.res.rr.com) has joined #ceph
[8:19] * topro (~topro@host-62-245-142-50.customer.m-online.net) has joined #ceph
[8:21] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) has joined #ceph
[8:22] * tnt (~tnt@91.176.35.47) has joined #ceph
[8:29] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) Quit (Ping timeout: 480 seconds)
[8:30] * sjusthm (~sam@24-205-35-233.dhcp.gldl.ca.charter.com) Quit (Ping timeout: 480 seconds)
[8:36] <huangjun> and my kernel is 3.8.13
[8:37] <huangjun> can not succeed on deploy osd on it
[8:40] * Vjarjadian (~IceChat77@176.254.37.210) Quit (Quit: A day without sunshine is like .... night)
[8:40] * tziOm (~bjornar@194.19.106.242) has joined #ceph
[8:41] * JustEra (~JustEra@ALille-555-1-227-212.w86-215.abo.wanadoo.fr) Quit (Quit: This computer has gone to sleep)
[8:45] * MarkN1 (~nathan@142.208.70.115.static.exetel.com.au) has joined #ceph
[8:45] * MarkN1 (~nathan@142.208.70.115.static.exetel.com.au) Quit ()
[8:47] * MarkN1 (~nathan@142.208.70.115.static.exetel.com.au) has joined #ceph
[8:47] * MarkN1 (~nathan@142.208.70.115.static.exetel.com.au) Quit ()
[8:49] * AfC1 (~andrew@2407:7800:200:1011:2ad2:44ff:fe08:a4c) Quit (Quit: Leaving.)
[8:54] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) Quit (Ping timeout: 480 seconds)
[8:55] * odyssey4me (~odyssey4m@165.233.71.2) has joined #ceph
[9:04] * BManojlovic (~steki@91.195.39.5) has joined #ceph
[9:07] * JustEra (~JustEra@89.234.148.11) has joined #ceph
[9:09] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) has joined #ceph
[9:10] * wogri_risc (~wogri_ris@ro.risc.uni-linz.ac.at) has joined #ceph
[9:15] * yy-nm (~Thunderbi@125.122.52.139) Quit (Remote host closed the connection)
[9:16] * yy-nm (~Thunderbi@125.122.52.139) has joined #ceph
[9:25] * buck (~buck@c-24-6-91-4.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[9:27] * tnt (~tnt@91.176.35.47) Quit (Read error: Operation timed out)
[9:27] * Cube (~Cube@cpe-76-95-217-129.socal.res.rr.com) Quit (Quit: Leaving.)
[9:36] * Bada (~Bada@195.65.225.142) has joined #ceph
[9:36] * ptosch (~root@82.113.99.65) has joined #ceph
[9:37] * ptosch (~root@82.113.99.65) Quit ()
[9:39] * ScOut3R (~ScOut3R@catv-89-133-25-52.catv.broadband.hu) has joined #ceph
[9:45] * DarkAceZ (~BillyMays@50.107.55.36) Quit (Ping timeout: 480 seconds)
[9:47] * tnt (~tnt@212-166-48-236.win.be) has joined #ceph
[10:01] * DarkAceZ (~BillyMays@50.107.55.36) has joined #ceph
[10:02] * sleinen1 (~Adium@2001:620:0:25:3960:8a8:ade6:480a) has joined #ceph
[10:09] * jcsp (~john@82-71-55-202.dsl.in-addr.zen.co.uk) has joined #ceph
[10:10] * jbd_ (~jbd_@2001:41d0:52:a00::77) has joined #ceph
[10:12] * LeaChim (~LeaChim@31.104.108.205) has joined #ceph
[10:21] * LeaChim (~LeaChim@31.104.108.205) Quit (Ping timeout: 480 seconds)
[10:28] * allsystemsarego (~allsystem@188.27.167.103) has joined #ceph
[10:30] * pieter_ (~pieter@105-236-69-62.access.mtnbusiness.co.za) has joined #ceph
[10:36] * LeaChim (~LeaChim@97e00ac2.skybroadband.com) has joined #ceph
[10:37] <pieter_> I have 2x 2TB OSD's, which I want to expose via CephFS. I'd like them to work as a sort of network RAID1, and so how do I get the size to show as 2TB instead of 4TB after mounting?
[10:41] <tnt> you can't
[10:41] * markl (~mark@tpsit.com) Quit (Read error: Operation timed out)
[10:41] <nerdtron> pieter_, ceph is more like "replicate" and not raid
[10:43] <tnt> and because you could actually ask for more or less than 2 replica even per-directory, ceph won't display 2G. It will always display the raw space and then whenever some gets used, it will be deducted in function on how many replicas were done.
[10:44] <pieter_> ok, cool - thanks
[10:46] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) Quit (Quit: tryggvil)
[10:48] * Isaac_ (~ijh@130.216.30.112) has joined #ceph
[11:03] <nerdtron> pieter_, ceph by default is set to replicate 2 times... say you save a 5GB file, it will occupy 10GB, the copy is saved/distributed on the nodes depending on the CRUSH algorithm
[11:13] * smiley_ (~smiley@pool-173-73-0-53.washdc.fios.verizon.net) Quit (Quit: smiley_)
[11:18] <pieter_> nerdtron: thanks, discovered that. so if one of my machines/osd's dies, and given that I have min-size = 1, IO will still work, right?
[11:18] * yy-nm (~Thunderbi@125.122.52.139) Quit (Quit: yy-nm)
[11:18] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[11:19] <nerdtron> yes, but how many monitors do you have? if only the osd dies, but the mon is still alive, you data is still good,
[11:19] <nerdtron> you can still read/write data
[11:19] * KindOne (~KindOne@0001a7db.user.oftc.net) has joined #ceph
[11:23] <wogri_risc> pieter_ - as long as a majority of your mon's is up (more than half) and any osd still has a replica of your data ceph will work
[11:24] * themgt_ (~themgt@201-223-248-202.baf.movistar.cl) has joined #ceph
[11:30] * themgt (~themgt@201-223-217-113.baf.movistar.cl) Quit (Ping timeout: 480 seconds)
[11:30] * themgt_ is now known as themgt
[11:31] * thomnico (~thomnico@2a01:e35:8b41:120:dc85:9784:3cef:7288) has joined #ceph
[11:40] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) has joined #ceph
[11:43] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) Quit (Remote host closed the connection)
[11:43] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) has joined #ceph
[11:47] * tryggvil (~tryggvil@178.19.53.254) has joined #ceph
[12:00] <pieter_> got it, thanks guys
[12:00] * thomnico (~thomnico@2a01:e35:8b41:120:dc85:9784:3cef:7288) Quit (Read error: Connection reset by peer)
[12:03] * yanzheng (~zhyan@jfdmzpr01-ext.jf.intel.com) Quit (Remote host closed the connection)
[12:10] * Isaac_ (~ijh@130.216.30.112) Quit (Remote host closed the connection)
[12:39] * hybrid512 (~walid@LPoitiers-156-86-25-85.w193-248.abo.wanadoo.fr) Quit (Read error: Connection reset by peer)
[12:41] * thomnico (~thomnico@2a01:e35:8b41:120:3873:745d:4b33:d33) has joined #ceph
[12:54] * freedomhui (~freedomhu@106.3.103.180) Quit (Quit: Leaving...)
[12:55] * mattt (~mattt@94.236.7.190) has joined #ceph
[13:00] * huangjun (~kvirc@111.172.32.75) Quit (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/)
[13:01] * nerdtron (~kenneth@202.60.8.252) Quit (Quit: Leaving)
[13:11] * ScOut3R (~ScOut3R@catv-89-133-25-52.catv.broadband.hu) Quit (Ping timeout: 480 seconds)
[13:18] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) has joined #ceph
[13:26] * mtk (~mtk@ool-44c35983.dyn.optonline.net) Quit (Remote host closed the connection)
[13:28] * KevinPerks (~Adium@cpe-066-026-239-136.triad.res.rr.com) has joined #ceph
[13:28] * Kioob (~kioob@2a01:e35:2432:58a0:21e:8cff:fe07:45b6) Quit (Ping timeout: 480 seconds)
[13:34] * smiley_ (~smiley@pool-173-73-0-53.washdc.fios.verizon.net) has joined #ceph
[13:36] * diegows (~diegows@190.190.11.42) has joined #ceph
[13:37] * wogri_risc (~wogri_ris@ro.risc.uni-linz.ac.at) Quit (Ping timeout: 480 seconds)
[13:41] * cofol1986 (~xwrj@110.90.119.113) has joined #ceph
[13:43] <cofol1986> The status of cephfs is health, but can't "ls " the dir, but can ls the file under the dir.
[13:43] <cofol1986> can anyone give some advices?
[13:53] * thomnico (~thomnico@2a01:e35:8b41:120:3873:745d:4b33:d33) Quit (Ping timeout: 480 seconds)
[13:54] * ScOut3R (~ScOut3R@catv-89-133-17-71.catv.broadband.hu) has joined #ceph
[13:58] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[13:58] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[14:01] * thomnico (~thomnico@2a01:e35:8b41:120:dc85:9784:3cef:7288) has joined #ceph
[14:05] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) has joined #ceph
[14:05] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) has joined #ceph
[14:07] <pieter_> any ideas on how large the journal has to be? is there a rule vs the size of the osd?
[14:08] * jlogan2 (~Thunderbi@2600:c00:3010:1:1::40) Quit (Ping timeout: 480 seconds)
[14:08] * thorus (~jonas@82.199.158.66) has joined #ceph
[14:09] * hybrid512 (~walid@LPoitiers-156-86-25-85.w193-248.abo.wanadoo.fr) has joined #ceph
[14:10] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) Quit (Ping timeout: 480 seconds)
[14:14] <thorus> I'm on cuttlefish (0.61.8), I have 4 nodes, 3 with 10 osd and one with 35 osds, I have a public network and a cluster network(both 1Gb/s) The one Node with 35 osds is not fully in yet because every time I add an osd it has an hugh impact on the read and write performance of the cluster. Is there a good practice way to add a new osd to minimalize the impact?
[14:17] * smiley_ (~smiley@pool-173-73-0-53.washdc.fios.verizon.net) Quit (Quit: smiley_)
[14:20] * cofol1986 (~xwrj@110.90.119.113) Quit (Read error: Connection reset by peer)
[14:21] * freedomhui (~freedomhu@117.79.232.216) has joined #ceph
[14:33] * jbd_ (~jbd_@2001:41d0:52:a00::77) Quit (Read error: No route to host)
[14:33] * jbd_ (~jbd_@2001:41d0:52:a00::77) has joined #ceph
[14:34] * mtk (~mtk@ool-44c35983.dyn.optonline.net) has joined #ceph
[14:35] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) Quit (Quit: Leaving.)
[14:36] * lx0 (~aoliva@lxo.user.oftc.net) has joined #ceph
[14:36] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) has joined #ceph
[14:36] * markbby (~Adium@168.94.245.1) has joined #ceph
[14:39] * imjustmatthew (~imjustmat@c-24-127-107-51.hsd1.va.comcast.net) has joined #ceph
[14:39] * foosinn (~stefan@office.unitedcolo.de) has joined #ceph
[14:41] * lxo (~aoliva@lxo.user.oftc.net) Quit (Ping timeout: 480 seconds)
[14:44] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) has joined #ceph
[14:45] * smiley_ (~smiley@pool-173-73-0-53.washdc.fios.verizon.net) has joined #ceph
[14:54] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) Quit (Ping timeout: 480 seconds)
[14:54] * smiley_ (~smiley@pool-173-73-0-53.washdc.fios.verizon.net) Quit (Quit: smiley_)
[14:55] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) has joined #ceph
[14:59] * julian_ (~julianwa@125.70.133.115) has joined #ceph
[15:03] * clayb (~kvirc@proxy-ny2.bloomberg.com) has joined #ceph
[15:03] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) has joined #ceph
[15:03] * clayb (~kvirc@proxy-ny2.bloomberg.com) Quit ()
[15:03] * clayb (~kvirc@proxy-ny2.bloomberg.com) has joined #ceph
[15:04] * julian (~julianwa@125.70.133.115) Quit (Ping timeout: 480 seconds)
[15:14] * allsystemsarego (~allsystem@188.27.167.103) Quit (Quit: Leaving)
[15:25] * sprachgenerator (~sprachgen@c-50-141-192-36.hsd1.il.comcast.net) has joined #ceph
[15:30] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[15:30] * ChanServ sets mode +v andreask
[15:34] * sprachgenerator (~sprachgen@c-50-141-192-36.hsd1.il.comcast.net) Quit (Quit: sprachgenerator)
[15:35] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[15:41] * dmsimard (~Adium@ap02.wireless.co.mtl.iweb.com) has joined #ceph
[15:43] <yo61> Afternoon
[15:44] <joelio> 'lo
[15:46] <yo61> I'm using ceph as an openstack backedn
[15:46] <yo61> ie. for volumes
[15:46] <yo61> It seems to "work" in that I can create volumes and boot off them
[15:47] <yo61> But I'm finding that some operations "hang" in the instance when accessing the disk
[15:47] <yo61> I can't figure out why
[15:47] <yo61> I can echo "stuff" > test.txt
[15:47] <yo61> and rm -f test.txt
[15:47] <yo61> But if I vi test.txt
[15:48] <yo61> then the process hangs when I write the file
[15:49] * wogri_risc (~wogri_ris@ro.risc.uni-linz.ac.at) has joined #ceph
[15:49] * wogri_risc (~wogri_ris@ro.risc.uni-linz.ac.at) Quit ()
[15:49] * wogri_risc (~wogri_ris@ro.risc.uni-linz.ac.at) has joined #ceph
[15:50] <yo61> Also, I can scp an rpm to the instance but rpm -Uvh hangs
[15:53] * wogri_risc (~wogri_ris@ro.risc.uni-linz.ac.at) Quit (Remote host closed the connection)
[15:55] * markbby (~Adium@168.94.245.1) Quit (Quit: Leaving.)
[15:56] * yanzheng (~zhyan@jfdmzpr05-ext.jf.intel.com) has joined #ceph
[15:56] * BillK (~BillK-OFT@58-7-131-166.dyn.iinet.net.au) Quit (Ping timeout: 480 seconds)
[15:58] * skm (~smiley@205.153.36.170) Quit (Read error: Connection reset by peer)
[15:58] <yo61> Most odd
[16:00] * skm (~smiley@205.153.36.170) has joined #ceph
[16:04] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) Quit (Quit: Leaving.)
[16:04] * markl (~mark@tpsit.com) has joined #ceph
[16:04] * markl (~mark@tpsit.com) Quit ()
[16:05] * markl (~mark@tpsit.com) has joined #ceph
[16:06] <decede> yo61: worth an strace to see what its hanging on in the instance?
[16:08] <yo61> strace not installed and can't install it
[16:08] <yo61> (that's the rpm I copied to the instance and tried to install :) )
[16:08] * Kioob`Taff (~plug-oliv@local.plusdinfo.com) has joined #ceph
[16:09] <dmsimard> If I'm correct in the assumption that writes are done properly (insert string to file, delete file, scp a file), it seems you have a issue with reads (opening file with vi, installing a package with rpm)
[16:09] <skm> stage1-mysql
[16:09] <dmsimard> I don't have a solution but you should probably look in that direction
[16:09] <yo61> I can open the file OK with vi
[16:09] <yo61> And cat it
[16:09] <dmsimard> But you said it hang ?
[16:10] <yo61> It's when I try and write
[16:10] <yo61> (that it hangs)
[16:10] <dmsimard> oh, when you save the file
[16:10] <dmsimard> Do you have hdparm ?à
[16:10] <dmsimard> the package/binary
[16:11] <yo61> No, not installed
[16:11] <yanzheng> cat /proc/<pid>/stack
[16:12] <yanzheng> you will see hang task's call stack
[16:13] <yo61> https://gist.github.com/robinbowes/6437506
[16:15] * jlogan1 (~Thunderbi@2600:c00:3010:1:1::40) has joined #ceph
[16:15] <yo61> OK, got hdparm on the box
[16:15] * markbby (~Adium@168.94.245.1) has joined #ceph
[16:15] <yo61> (copied the binary from another install)
[16:16] <yanzheng> are there errors in the kernel message
[16:16] <pieter_> does it make sense to mount RDB and expose via NFS to web servers for shared data vs 'unstable' cephfs?
[16:16] <pieter_> *RBD
[16:16] <yo61> yanzheng: where?
[16:17] <yanzheng> dmesg
[16:17] <pieter_> or can more than one client mount the same RBD?
[16:17] <yo61> I don't see any
[16:18] <yanzheng> cat /sys/kernel/debug/ceph/*/osdc
[16:18] <Gugge-47527> pieter_: you cant mount an RBD
[16:18] <yo61> yanzheng: on the hypervisor?
[16:19] <Gugge-47527> pieter_: but you can map an RBD on multiple machines at once, and use a clusterfilesystem on top
[16:19] <yanzheng> you are not using kernel rbd?
[16:20] <yo61> I am using ceph rbd via libvirt
[16:20] <pieter_> Ahh, thanks Gugge-47527 I see I misunderstood. Makes sense now
[16:21] * tryggvil (~tryggvil@178.19.53.254) Quit (Quit: tryggvil)
[16:21] <pieter_> I think I will try cephfs for my 2x web servers for common images
[16:21] <pieter_> shared images*
[16:22] <pieter_> in my vm's it seems stable enough...
[16:22] <yo61> It seems to boot OK, and writes logs OK, but some operations are hanging
[16:23] * dosaboy_ (~dosaboy@host86-156-255-54.range86-156.btcentralplus.com) has joined #ceph
[16:29] * dosaboy (~dosaboy@host81-155-236-130.range81-155.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[16:31] * jeff-YF (~jeffyf@67.23.117.122) has joined #ceph
[16:32] * tziOm (~bjornar@194.19.106.242) Quit (Remote host closed the connection)
[16:32] * jeff-YF (~jeffyf@67.23.117.122) Quit (Remote host closed the connection)
[16:32] * jeff-YF (~jeffyf@67.23.117.122) has joined #ceph
[16:32] * julian_ (~julianwa@125.70.133.115) Quit (Quit: afk)
[16:32] * dosaboy (~dosaboy@host109-155-27-244.range109-155.btcentralplus.com) has joined #ceph
[16:37] * PerlStalker (~PerlStalk@2620:d3:8000:192::70) has joined #ceph
[16:37] * dmsimard1 (~Adium@108.163.152.2) has joined #ceph
[16:37] * dosaboy__ (~dosaboy@host109-155-13-93.range109-155.btcentralplus.com) has joined #ceph
[16:39] * dosaboy_ (~dosaboy@host86-156-255-54.range86-156.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[16:39] * dmsimard (~Adium@ap02.wireless.co.mtl.iweb.com) Quit (Read error: Connection reset by peer)
[16:41] <joelio> yo61: have you got rbd caching enabled in config but not set cache in libvirt config?
[16:42] * dosaboy (~dosaboy@host109-155-27-244.range109-155.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[16:44] <yo61> Hmm, idk
[16:44] <yo61> How do I test that?
[16:44] <yo61> (check that)
[16:44] * mikedawson_ (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) has joined #ceph
[16:44] * yanzheng (~zhyan@jfdmzpr05-ext.jf.intel.com) Quit (Ping timeout: 480 seconds)
[16:45] <yo61> I haven't conciously set cache on or off in either place
[16:48] <LCF> https://github.com/ceph/ceph/pull/564 <- I should do something more beside pull request if I notice bug ?
[16:48] <yo61> joelio: ^^
[16:51] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) Quit (Ping timeout: 480 seconds)
[16:51] * mikedawson_ is now known as mikedawson
[16:51] <joelio> yo61: you can check the status of rbd caching (and other stuff) via an admin socket
[16:52] <joelio> no an openstack user, but if you have it enabled, then you need to ensure a cache strategy (writeback/throuh) is set too in the vm's xml
[16:53] <joelio> as you're suffering on writes, seems like it could be a write cache issue - but I'm not 100% to be honest
[16:54] <yehudasa_> ccourtaut: around?
[16:54] <yo61> OK, thanks
[16:54] <ccourtaut> yehudasa_: yes
[16:55] <ccourtaut> how are you doing?
[16:55] <joelio> yo61: http://www.sebastien-han.fr/blog/2012/08/14/ceph-admin-socket/
[16:55] <yehudasa_> ccourtaut: great, I understand you look at the sync-agent nowadays?
[16:55] <ccourtaut> yehudasa_: yes actually i am
[16:55] <ccourtaut> i've seen the progress here
[16:56] <ccourtaut> but i was wondering what was going next, and of course, how
[16:56] <yehudasa_> ccourtaut: the next on the plate is doing the data sync
[16:56] <ccourtaut> i've seen data replication just appeared, but, it uses what seems to be inter-region copy, and i do not see why
[16:57] <ccourtaut> yehudasa_: yes that was my conclusion
[16:57] <ccourtaut> what i am trying to figure out is the correct way to do it
[16:57] <yehudasa_> well, it should use intra-region copy
[16:57] <yehudasa_> according to the current design at least
[16:57] <yehudasa_> basically the agents send copy requests for objects that need to be synce
[16:57] <yehudasa_> syncde
[16:58] <ccourtaut> intra-region isn't meant for use to copy data from a master zone to a slave zone?
[16:58] <yehudasa_> synced.. for the intra-region copy (when you copy objects within the same region from different zones) the object attributes are preserved
[16:59] <yehudasa_> .. it is
[16:59] <yehudasa_> I might be missing a point here
[17:00] <ccourtaut> yehudasa_: is this the same api call to sync from a master-region/master-zone to a slave-region/master-zone
[17:01] <yehudasa_> it's similar to the call to copy objects between regions
[17:01] <yehudasa_> the params might be a bit different
[17:01] <yehudasa_> http://wiki.ceph.com/index.php?title=RESTful_API_for_DR_%2F%2F_Geo-Replication
[17:01] * ishkabob (~c7a82cc0@webuser.thegrebs.com) has joined #ceph
[17:01] <ntranger> I was just asked by a buddy here at work, he wants to setup ceph at his house, for his personal stuff. He asked which is the ideal linux distro for ceph?
[17:02] <ccourtaut> yehudasa_: you're referring to this API call to do inter-region copy? Fetch object from region (inter-region copy)
[17:02] <yehudasa_> there's the intra-region copy
[17:03] <ishkabob> hey guys, does anyone know why i'm having trouble deleting a pool? http://pastebin.com/NmAUrZs0
[17:03] <ccourtaut> the intra region copy is the currently implemented method for data sync in the wip-buck-datasync branch of radosgw-agent?
[17:03] * pieter_ (~pieter@105-236-69-62.access.mtnbusiness.co.za) Quit (Quit: Konversation terminated!)
[17:04] <ishkabob> ps - i was able to successfully delete it with the rados command
[17:04] <yehudasa_> ccourtaut: I haven't looked at that implementation yet, that what at least should be used. I know there was some mixup there a few days ago
[17:05] <ccourtaut> ok
[17:06] <ccourtaut> yehudasa_: i'm just trying to figure what needs to be done, and with which tools that are already present in radosgw
[17:07] * JustEra (~JustEra@89.234.148.11) Quit (Ping timeout: 480 seconds)
[17:08] <yehudasa_> ccourtaut: I created a few issues on the tracker that roughly split the work into consumable tasks (6190-6195)
[17:09] <ccourtaut> oh ok
[17:09] <ccourtaut> great
[17:09] <yehudasa_> joshd is also supposed to take over jbuck's work
[17:09] * ccourtaut opening the tracker
[17:09] * Vjarjadian (~IceChat77@176.254.37.210) has joined #ceph
[17:09] <yehudasa_> .. once he finds some time out of rbd stuff
[17:09] <ccourtaut> yehudasa_: ok, so i should stay in touch with joshd for the future work?
[17:10] <yehudasa_> maybe stay in touch with both of us? I'm trying to move it forward
[17:10] <ccourtaut> yehudasa_: yep sure
[17:10] <yehudasa_> there are a few rough edges and maybe missing some apis
[17:11] <yehudasa_> it'd be great if someone could do some manual walk through the sync process
[17:11] <ccourtaut> yehudasa_: ok, feel free to ping me and assign tasks, i will be glad to do this as far as my understanding can lead me
[17:12] <yehudasa_> also, creating tests tasks is always helpful
[17:12] <ccourtaut> yehudasa_: i'm not sure to understand the "manual walk through"
[17:12] <ccourtaut> yehudasa_: tests tasks, you mean in teuthology
[17:13] <yehudasa_> e.g., set up two zones within a single region, call manually the needed apis in order to get the sync done
[17:13] <yehudasa_> yeah, teuthology
[17:14] * BManojlovic (~steki@91.195.39.5) Quit (Quit: Ja odoh a vi sta 'ocete...)
[17:14] <yehudasa_> ccourtaut: also if you want and have the time for it you can join our standup
[17:14] <ccourtaut> ok, i'll try to do that, as i got a little script to setup a local cluster, i will try to modify it a bit to have a full setup with multiple region/zone, so that i'll be able to test intra/inter region sync
[17:14] <yehudasa_> it's in 9:45 pacific
[17:15] <ntranger> hey, alfredodeza, what linux distro would you recommend for ceph?
[17:15] <alfredodeza> ntranger: we support quite a few
[17:15] <alfredodeza> I don't think we push towards one distro
[17:15] <alfredodeza> CentOS, Ubuntu, RedHat, Fedora are a few
[17:16] <ccourtaut> yehudasa_: i'll take a look to my schedule to see if i'll be able to join the standup
[17:16] <ccourtaut> but would be glad to
[17:16] <ccourtaut> we keep in touch about that
[17:16] <yehudasa_> ccourtaut: sure, I can send you a hangout invite, you can join if you want to
[17:16] <ccourtaut> yep sure, but it would be more if i can rather than want :)
[17:18] <yehudasa_> ccourtaut: sure
[17:18] <yehudasa_> sent invitation (I think)
[17:20] <ccourtaut> yehudasa_: didn't received anything
[17:21] <yehudasa_> ccourtaut: it was supposed to go to your gmail, anything there?
[17:22] <ccourtaut> yehudasa_: that is where i looked first (i thought that hangouts would need gmail address)
[17:22] <ccourtaut> yehudasa_: maybe gmail is jetlagged
[17:24] * tryggvil (~tryggvil@178.19.53.254) has joined #ceph
[17:25] * rovar (~oftc-webi@69.191.241.59) has joined #ceph
[17:26] <rovar> quick question: I'm running 0.67.2 and I'm simply trying to delete a pool
[17:26] <rovar> ceph osd pool delete pbench
[17:26] <rovar> has the command line format changed?
[17:27] * rovar (~oftc-webi@69.191.241.59) Quit (Remote host closed the connection)
[17:29] * Bada (~Bada@195.65.225.142) Quit (Ping timeout: 480 seconds)
[17:31] * egyptkjbtyyi4yt9845jsks (~egyptmosl@41.153.202.238) has joined #ceph
[17:31] * egyptkjbtyyi4yt9845jsks (~egyptmosl@41.153.202.238) Quit (autokilled: Do not spam. Mail support@oftc.net with questions. (2013-09-04 15:31:45))
[17:32] * tryggvil (~tryggvil@178.19.53.254) Quit (Quit: tryggvil)
[17:32] * The_Bishop (~bishop@f055002034.adsl.alicedsl.de) has joined #ceph
[17:32] * rovar (~oftc-webi@proxy-nj1.bloomberg.com) has joined #ceph
[17:32] <rovar> argh
[17:32] * sagelap (~sage@2600:1012:b002:2ff:a1d8:6962:87a3:483b) has joined #ceph
[17:32] <rovar> ok.. if anyone responded to my last query, sorry I missed it.
[17:34] * tryggvil (~tryggvil@178.19.53.254) has joined #ceph
[17:34] * vata (~vata@2607:fad8:4:6:b93b:c4fe:b6d0:d484) has joined #ceph
[17:34] <yehudasa_> rovar: it might have, I'm not sure, but you can use the rados command instead (rados rmpool)
[17:35] <rovar> ah.. I think I see the problem..
[17:36] <rovar> yehudasa_: the problem was that of an unhelpful message
[17:36] <rovar> running rados rmpool tells me that I need to pass the pool name twice plus --yes...
[17:36] <rovar> so running the osd pool delete command in that way worked.
[17:37] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) has joined #ceph
[17:47] * freedomhui (~freedomhu@117.79.232.216) Quit (Quit: Leaving...)
[17:47] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Ping timeout: 480 seconds)
[17:48] * bandrus (~Adium@cpe-76-95-217-129.socal.res.rr.com) has joined #ceph
[17:48] * ross_ (~ross@60.208.111.209) has joined #ceph
[17:50] * sprachgenerator (~sprachgen@130.202.135.214) has joined #ceph
[17:54] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[17:54] * ChanServ sets mode +v andreask
[17:55] * ross__ (~ross@60.208.111.209) Quit (Ping timeout: 480 seconds)
[17:59] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) has joined #ceph
[18:01] * gregmark (~Adium@68.87.42.115) has joined #ceph
[18:01] * xarses (~andreww@c-71-202-167-197.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[18:02] * foosinn (~stefan@office.unitedcolo.de) Quit (Quit: Leaving)
[18:05] * ScOut3R (~ScOut3R@catv-89-133-17-71.catv.broadband.hu) Quit (Ping timeout: 480 seconds)
[18:08] * mattt (~mattt@94.236.7.190) Quit (Quit: Computer has gone to sleep.)
[18:09] * markbby (~Adium@168.94.245.1) Quit (Remote host closed the connection)
[18:09] * markbby (~Adium@168.94.245.1) has joined #ceph
[18:10] * nwat (~nwat@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[18:10] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) has joined #ceph
[18:12] <sagewk> zackc: a couple small teuthology tweaks, when you have a minute: https://github.com/organizations/ceph/dashboard/pulls/for/ceph/teuthology
[18:13] * sagelap (~sage@2600:1012:b002:2ff:a1d8:6962:87a3:483b) Quit (Ping timeout: 480 seconds)
[18:15] * odyssey4me (~odyssey4m@165.233.71.2) Quit (Ping timeout: 480 seconds)
[18:16] <zackc> sagewk: merged!
[18:16] <sagewk> thanks!
[18:17] * tnt (~tnt@212-166-48-236.win.be) Quit (Ping timeout: 480 seconds)
[18:17] * angdraug (~angdraug@204.11.231.50.static.etheric.net) has joined #ceph
[18:23] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) Quit (Quit: This computer has gone to sleep)
[18:26] * haomaiwang (~haomaiwan@112.193.130.196) Quit (Ping timeout: 480 seconds)
[18:26] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) has joined #ceph
[18:27] * The_Bishop_ (~bishop@g227137157.adsl.alicedsl.de) has joined #ceph
[18:27] * dosaboy__ is now known as dosaboy
[18:28] * The_Bishop (~bishop@f055002034.adsl.alicedsl.de) Quit (Read error: Operation timed out)
[18:29] * sagelap (~sage@2607:f298:a:607:ea03:9aff:febc:4c23) has joined #ceph
[18:29] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Ping timeout: 480 seconds)
[18:31] * ishkabob (~c7a82cc0@webuser.thegrebs.com) Quit (Quit: TheGrebs.com CGI:IRC (Ping timeout))
[18:32] * gregaf1 (~Adium@cpe-172-250-66-79.socal.res.rr.com) has joined #ceph
[18:35] * xarses (~andreww@204.11.231.50.static.etheric.net) has joined #ceph
[18:35] * nwat (~nwat@eduroam-237-79.ucsc.edu) has joined #ceph
[18:35] * marrusl_ (~mark@209-150-43-182.c3-0.wsd-ubr2.qens-wsd.ny.cable.rcn.com) has joined #ceph
[18:35] * marrusl (~mark@209-150-43-182.c3-0.wsd-ubr2.qens-wsd.ny.cable.rcn.com) Quit (Read error: Connection reset by peer)
[18:37] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) Quit (Quit: This computer has gone to sleep)
[18:40] * tnt (~tnt@91.176.35.47) has joined #ceph
[18:45] * lx0 (~aoliva@lxo.user.oftc.net) Quit (Remote host closed the connection)
[18:45] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) has joined #ceph
[18:47] * lx0 (~aoliva@lxo.user.oftc.net) has joined #ceph
[18:48] * alram (~alram@38.122.20.226) has joined #ceph
[18:50] <rsanti> about http://ceph.com/docs/next/rados/operations/authentication/#enabling-cephx - do I need to worry about distributing those files if I don't use ceph-deploy?
[18:53] * jeff-YF (~jeffyf@67.23.117.122) Quit (Quit: jeff-YF)
[18:54] <rsanti> I mean, beside the monitor keyring
[18:54] <rsanti> that's spelled out in the guide
[18:57] * eternaleye (~eternaley@2002:3284:29cb::1) Quit (Ping timeout: 480 seconds)
[18:58] * eternaleye (~eternaley@2002:3284:29cb::1) has joined #ceph
[18:59] <sagewk> rsanti: which files?
[18:59] <rovar> rsanti, I think you do. We use chef to spin up 1 mon node at a time
[18:59] <rovar> ceph.conf, I'm assuming
[19:00] <rsanti> e.g. /var/lib/ceph/osd/ceph-{$id}/keyring but also ceph.conf actually.
[19:00] <rovar> sagewk: does rados bench do some sort of locking on a pool? I'm trying to run a write bench on a pool and a read bench from another machine on the same pool and I'm getting -2 error
[19:00] * Vjarjadian (~IceChat77@176.254.37.210) Quit (Quit: Copywight 2007 Elmer Fudd. All wights wesewved.)
[19:01] <rovar> rsanti: yea. The presence of the directories under /var/lib/ceph/... are what tells ceph that those osd's are present, so that you don't have to enumerate them in the .conf
[19:02] * marrusl_ is now known as marrusl
[19:03] * rovar (~oftc-webi@proxy-nj1.bloomberg.com) Quit (Remote host closed the connection)
[19:04] <PerlStalker> Is there a way to lower the priority or resource usage for backfills besides the 'osd backfill *' settings?
[19:05] <PerlStalker> I replaced a drive yesterday and the backfills flooded my network and severely hurt my cluster performance.
[19:07] * jeff-YF (~jeffyf@67.23.117.122) has joined #ceph
[19:07] * ScOut3R (~scout3r@BC2484D1.dsl.pool.telekom.hu) has joined #ceph
[19:13] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) Quit (Quit: This computer has gone to sleep)
[19:17] * rovar (~oftc-webi@proxy-nj1.bloomberg.com) has joined #ceph
[19:17] * yehudasa_ (~yehudasa@2602:306:330b:1410:3c66:2012:30e2:5efb) Quit (Remote host closed the connection)
[19:25] * Cube (~Cube@cpe-76-95-217-129.socal.res.rr.com) has joined #ceph
[19:26] * peetaur (~peter@CPEbc1401e60493-CMbc1401e60490.cpe.net.cable.rogers.com) Quit (Ping timeout: 480 seconds)
[19:30] * rovar (~oftc-webi@proxy-nj1.bloomberg.com) Quit (Remote host closed the connection)
[19:31] * sjusthm (~sam@24.205.35.233) has joined #ceph
[19:32] <PerlStalker> On a related note: How much of a benefit is there from splitting off the public and cluster networks?
[19:32] <PerlStalker> Is there a defined way of moving to separate networks without breaking the cluster?
[19:33] <dmsimard1> PerlStalker: The documentation mentions especially security
[19:34] <Gugge-47527> setup the networks on all the osd nodes, then chance the config, and restart the osd's
[19:34] <Gugge-47527> change
[19:34] <dmsimard1> The cluster network can be natively used between OSD nodes to replicate data
[19:34] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Remote host closed the connection)
[19:35] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[19:35] * LPG (~LPG@c-76-104-197-224.hsd1.wa.comcast.net) Quit (Remote host closed the connection)
[19:36] * rovar (~oftc-webi@69.191.241.59) has joined #ceph
[19:36] <PerlStalker> Gugge-47527: That's one of my questions. How will the cluster behave while some nodes are on the new network and others haven't made the move yet?
[19:36] <rovar> when one creates a user via radosgw-admin, do they automatically have perms to create buckets? I'm getting an ErrorAccessDenied from rados-bench
[19:37] * yehuda_hm (~yehuda@2602:306:330b:1410:baac:6fff:fec5:2aad) has joined #ceph
[19:39] * cjh_ (~cjh@ps123903.dreamhost.com) has joined #ceph
[19:42] * sleinen1 (~Adium@2001:620:0:25:3960:8a8:ade6:480a) Quit (Quit: Leaving.)
[19:42] * sleinen (~Adium@130.59.94.218) has joined #ceph
[19:45] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) Quit (Quit: Ex-Chat)
[19:50] <Gugge-47527> PerlStalker: if all nodes have the network setup, they can talk to each other between the networks
[19:50] <Gugge-47527> PerlStalker: and it will work
[19:50] * sleinen (~Adium@130.59.94.218) Quit (Ping timeout: 480 seconds)
[19:51] <PerlStalker> I see.
[19:51] <sjusthm> sagewk: that 6207 with logs appears to have failed at offset x which hadn't changed since the last read (which went fine) on the same osds
[19:52] <sjusthm> and kern.log had a btrfs backtrace
[19:52] <sagewk> ah!
[19:52] <sjusthm> so I am going to disregard that one for the moment
[19:52] <sagewk> k
[19:53] * JustEra (~JustEra@ALille-555-1-86-217.w90-7.abo.wanadoo.fr) has joined #ceph
[19:53] <sjusthm> sagewk: where was the one on dumpling?
[19:54] <sagewk> sjusthm: hmm, are you sure? i see an update before that read
[19:54] <sjusthm> oh, in the bug
[19:54] <sagewk> update_object_version oid 42 is version 3083
[19:54] <sjusthm> to different offsets
[19:55] <sjusthm> and there hadn't been recovery or replay
[19:56] * PerlStalker (~PerlStalk@2620:d3:8000:192::70) Quit (Remote host closed the connection)
[19:57] <sagewk> oh i see.
[19:57] <sagewk> i'll try to reproduce on xfs.
[19:57] <sjusthm> k
[19:57] <sjusthm> we've seen at least one example on ext4, so there is probably an actual bug
[19:57] <sjusthm> yeah, the dumpling one in the bug was on ext4
[20:00] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[20:08] <xarses> alfedodeza: you around?
[20:09] * alphe (~alphe@0001ac6f.user.oftc.net) has joined #ceph
[20:09] <xarses> alfredodeza even
[20:09] * jbd_ (~jbd_@2001:41d0:52:a00::77) has left #ceph
[20:09] <alphe> ceph-fuse 0.67-2 has a problem with kernel 3.8.0.19
[20:10] <alphe> Im downgrading to 0.61-8
[20:10] <alfredodeza> xarses: I am
[20:12] <xarses> for https://github.com/ceph/ceph-deploy/pull/58 I'm wondering why one shouldn't use a naked IP
[20:13] <xarses> when i use the dns name, it uses the wrong interface from public_network
[20:14] <xarses> i guess i should pass hostname:ip to force the correct interface?
[20:14] <ircolle> alphe - could you be more specific?
[20:15] <alphe> hum yes ... when I transfer files to ceph trhought samba or ftp suddently the transfer stops and from the gateway linux shell command line I can t enter the ceph mount point
[20:16] <alphe> this is what I notice I am not running ceph-fuse in debug mode
[20:16] * jluis is now known as joao
[20:16] * tziOm (~bjornar@ti0099a340-dhcp0395.bb.online.no) has joined #ceph
[20:16] <alphe> so I can t tell you precisely what is happenning
[20:16] <alfredodeza> xarses: using an IP (os opposed to NAME:IP) will have issues when creating the configuration file as we are splitting items on '.'
[20:16] <alfredodeza> xarses: for the correct interface you are right, you should explicitly do the pair
[20:17] <alphe> and that issue wasn t happenning with cuttle fish from 0.61-2 to 0.61-8
[20:17] <ircolle> alphe - ok. Have you already downgraded? Would you be willing to run in debug mode?
[20:17] * rturk-away is now known as rturk
[20:17] <alphe> ircolle I just uninstalled the ceph cluster but I could resinstall it I didn t purged it
[20:18] <ircolle> alphe - that would be a great help to get a look at what's going on
[20:18] <xarses> alfredodeza: OK, I'll have to fix my env.
[20:18] <alphe> ircolle :)
[20:18] <alphe> anything to help :)
[20:19] <alphe> I use regular kernel of the ubuntu 13.04 raring ...
[20:19] <alphe> and i already noticed that with that kernel the 0.67-2 cephFS was showing aberant sizing
[20:21] <alphe> ircolle hope that reinstall will not takes forever
[20:22] <alphe> hum ... strangely my ceph cluster is still working ...
[20:22] <alphe> probably because i didn t stoped it before doing the uninstall :)
[20:23] <alphe> ok so I can do the ceph-fuse debug mode how do i proceed ?
[20:25] <alphe> ircolle how do i proceed to get the debug with strace ?
[20:27] <ircolle> alphe - ceph-fuse —debug-client 20
[20:28] <alphe> then the -m osd01 /mount/dir?
[20:29] <alphe> or I mount then i do the debug-client call ?
[20:29] <alphe> I did it in one line
[20:30] <alphe> where do i look to get the results ?
[20:30] <ircolle> alphe - where is logging set up to?
[20:31] <alphe> hum for ceph fuse I don't know ...
[20:31] <alphe> in the ceph.conf I have log file = / var/log/ceph
[20:31] * grepory (~Adium@50-115-70-146.static-ip.telepacific.net) has joined #ceph
[20:32] <alphe> correction there is no log specified in my ceph.conf
[20:33] * jeff-YF (~jeffyf@67.23.117.122) Quit (Quit: jeff-YF)
[20:34] <alphe> 0.68 is out ...
[20:38] * peetaur (~peter@CPEbc1401e60493-CMbc1401e60490.cpe.net.cable.rogers.com) has joined #ceph
[20:38] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[20:38] * ChanServ sets mode +v andreask
[20:38] * jeff-YF (~jeffyf@67.23.123.228) has joined #ceph
[20:40] * tziOm (~bjornar@ti0099a340-dhcp0395.bb.online.no) Quit (Remote host closed the connection)
[20:41] * jeff-YF (~jeffyf@67.23.123.228) Quit ()
[20:43] <alphe> ircolle how do i enable loging ?
[20:43] <alphe> for ceph-fuse ?
[20:44] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) Quit (Ping timeout: 480 seconds)
[20:45] <ircolle> you can append —log-file=foo to previous command
[20:45] * sjm (~sjm@2607:f298:a:607:c5dd:1134:9d8a:e82f) has joined #ceph
[20:48] * tryggvil (~tryggvil@178.19.53.254) Quit (Quit: tryggvil)
[20:48] <alphe> ok
[20:50] <alphe> hehehehe tail -f on the log file is funny
[20:51] <alphe> OMG a gigazillion of text lines !
[20:51] <alphe> :))
[20:53] <alphe> the log file weights 150MB and increasing kids don t do that at home !
[20:53] <ircolle> alphe - yeah, level 20 is quite verbose
[20:54] <alphe> yep noticed that
[20:55] <alphe> until now there it is going smoothly
[20:56] * sprachgenerator (~sprachgen@130.202.135.214) Quit (Quit: sprachgenerator)
[20:56] <alphe> ok and freeezed
[20:57] <alphe> oh it resumed ...
[20:58] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) has joined #ceph
[20:59] * sprachgenerator (~sprachgen@130.202.135.214) has joined #ceph
[21:00] * PerlStalker (~PerlStalk@2620:d3:8000:192::70) has joined #ceph
[21:00] * marrusl (~mark@209-150-43-182.c3-0.wsd-ubr2.qens-wsd.ny.cable.rcn.com) Quit (Read error: Connection reset by peer)
[21:00] <sprachgenerator> has anyone had problems with the osd/mds and client admin key rings not being created when issuing a ceph-deploy mon create?
[21:01] <alphe> ok so now that the files transfer is stuck the only thing i see in the log is
[21:01] <alphe> client.5397 renew_caps mds.0
[21:01] <alphe> client.5397 handle_client_session client_session(renewcaps seq 36) v1 from mds.0
[21:01] <alphe> and time to time it increase
[21:01] <alphe> before that I got a !flush
[21:01] <alphe> don t know why
[21:01] * nhorman (~nhorman@hmsreliant.think-freely.org) has joined #ceph
[21:03] <alphe> ircolle do you want the whole log file ?
[21:04] * sjm (~sjm@2607:f298:a:607:c5dd:1134:9d8a:e82f) Quit (Remote host closed the connection)
[21:04] * tnt (~tnt@91.176.35.47) Quit (Ping timeout: 480 seconds)
[21:04] <ircolle> alphe - the whole thing
[21:04] <alfredodeza> sprachgenerator: sure
[21:04] <alfredodeza> there is a caveat
[21:04] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) has joined #ceph
[21:05] <alfredodeza> make sure the host you are providing to `mon create` matches the `hostname` in the remote host
[21:05] <alphe> ircolle one strange thing is that I had 2 ceph-fuse process ?
[21:05] <alfredodeza> for example, if you are doing `ceph-deploy mon create host1` then `host1` should also be the actual hostname in the remote
[21:05] <alphe> but I did umount before the second ceph-fuse call with the --log-file
[21:06] <alfredodeza> sprachgenerator: does that ring any bells?
[21:06] * sjm (~sjm@38.122.20.226) has joined #ceph
[21:06] * allsystemsarego (~allsystem@188.27.167.103) has joined #ceph
[21:07] <alphe> ircolle can you tell me where i can put that log file ?
[21:07] <alphe> server
[21:07] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[21:08] * rturk is now known as rturk-away
[21:08] * sjm (~sjm@38.122.20.226) has joined #ceph
[21:08] <sprachgenerator> I believe I am deploying the right way: http://pastebin.com/QG97cDGK
[21:09] * sagelap (~sage@2607:f298:a:607:ea03:9aff:febc:4c23) Quit (Quit: Leaving.)
[21:09] <ircolle> alphe - in a mtg - bbiab
[21:10] <alphe> uploading to filedroper is that ok ?
[21:10] <alphe> i will give you the link
[21:11] * rturk-away is now known as rturk
[21:13] <sprachgenerator> I noticed one inconsistency - I still fail in the same way - however I'm getting a few new steps in the ceph deploy : http://pastebin.com/wHeKgeWk
[21:16] <alphe> though bbiab was an big filesharing website >_<
[21:16] <alphe> :)
[21:16] * marrusl (~mark@209-150-43-182.c3-0.wsd-ubr2.qens-wsd.ny.cable.rcn.com) has joined #ceph
[21:17] <alphe> sprachgenerator [ceph01-p][INFO ] write cluster configuration to /etc/ceph/{cluster}.conf this dont smells good
[21:18] <alphe> previous to your mon creation did you do a ceph-deploy new clustername initial nodes ?
[21:18] <alphe> in ceph.conf you have initial members so that should be the case
[21:18] <alfredodeza> sprachgenerator: that doesn't tell me really if your remote hosts match the hostnames
[21:19] <alfredodeza> if you login to ceph01-p and call `hostname` what does it come back with?
[21:19] <alfredodeza> if it doesn't come back with ceph01-p, then that is your problem
[21:21] <sprachgenerator> here are copies of the configs for each node: http://pastebin.com/YFcFN91T - the hostname is @ the prompt for each respective entry - they match what is contained in the ceph.conf on each node
[21:21] <alphe> alfredodeza to downgrade an running ceph-cluster can I do ceph-deploy uninstall nodes then ceph-deploy install --stable cuttlefish nodes[allofthem]
[21:22] <alphe> sprachgenerator can you do a hostname from the shell commandline of each of your ceph-cluster nodes ?
[21:23] <sprachgenerator> alphe I did do this and it matches what is contained in ceph.conf
[21:24] <alphe> ok
[21:24] <sprachgenerator> ceph deploy is 1.2.3 and ceph version is 0.67.2
[21:26] <alphe> sprachgenerator when you created your monitors did you created them all at the same time ?
[21:27] <sprachgenerator> alphe yes I simply did a ceph-deploy new host1 host2 host3 and a ceph-deploy mon create host1 host2 host3
[21:27] <alphe> ok here is the prob
[21:27] <alphe> the new can be done on all your nodes that will be monitors
[21:27] <alphe> at the same time
[21:28] <alphe> but the monitor creation has to be with the first one then you pause for 30 secondes
[21:28] <alfredodeza> sprachgenerator: it looks like this might be something else
[21:28] <alphe> then you do the backup monitors
[21:28] <alfredodeza> have you tailed your mon logs?
[21:28] <alfredodeza> it will be better to increase the verbosity and restart the mons
[21:29] <alphe> alfredodeza I guess all the monitor are constantly calling for new elections ...
[21:29] <alfredodeza> it could be
[21:29] <alfredodeza> alas, that might be a bit out of my league :(
[21:30] <sprachgenerator> alpha I will try doing it with this approach and see where that leaves me, also I will grab some of the logs while I'm at it
[21:30] <alphe> this happends when you do a ceph-deploy mon create host1 host2 host3 on the same line
[21:30] <alphe> sprachgenerator that happends to me too what you need to do is purge all data
[21:31] <alfredodeza> sprachgenerator: in your ceph.conf, increase the verbosity in your global section with these lines:
[21:31] <alfredodeza> debug ms = 1
[21:31] <alfredodeza> debug mon = 20
[21:31] <alphe> then you stop ceph process
[21:31] <alfredodeza> oh, I think alphe is right
[21:32] * jeff-YF (~jeffyf@67.23.117.122) has joined #ceph
[21:32] <alphe> if you do ceph-deploy mon create host1 host2 host3 on the same line then all of your monitors will claim to be the master and no client.admin.keyring will be created ...
[21:33] * alfredodeza needs to update the docs on that
[21:33] <alphe> you have to create a first monitor that will create the admin keyring and the mon.keyring then you create the backups that will see the first monitor and retreive the keyrings they need
[21:33] <alphe> alfredodeza yes or put a sleep 30 seconds in the ceph-deploy script :)
[21:34] <alphe> for the iterations of the monitor creation cause everyone is intended to create all the monitors at the same time which is logical
[21:35] * LPG (~LPG@c-76-104-197-224.hsd1.wa.comcast.net) has joined #ceph
[21:36] * thomnico (~thomnico@2a01:e35:8b41:120:dc85:9784:3cef:7288) Quit (Ping timeout: 480 seconds)
[21:36] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) Quit (Quit: Ja odoh a vi sta 'ocete...)
[21:38] * jhujhiti (~jhujhiti@00012a8b.user.oftc.net) has joined #ceph
[21:38] <jhujhiti> is there any reason why i shouldn't run a dumpling client against a bobtail cluster?
[21:38] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) has joined #ceph
[21:40] * dosaboy_ (~dosaboy@host109-156-222-77.range109-156.btcentralplus.com) has joined #ceph
[21:40] <alphe> jhujhiti probably because they have like a zillion changes ...
[21:42] * grepory (~Adium@50-115-70-146.static-ip.telepacific.net) Quit (Quit: Leaving.)
[21:43] * Vjarjadian (~IceChat77@176.254.37.210) has joined #ceph
[21:43] * LPG (~LPG@c-76-104-197-224.hsd1.wa.comcast.net) Quit (Ping timeout: 480 seconds)
[21:44] <alphe> ircolle I think I found the problem for the ceph-fuse that freeze in 0.67-2
[21:44] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[21:44] <alphe> one of my node network output was jammed ...
[21:44] <ircolle> alphe - perfect timing, meeting just ended
[21:44] <alphe> it could receive data but not send the ack back
[21:45] <ircolle> ahh
[21:45] <alphe> http://www.filedropper.com/ceph-fuse
[21:45] <alphe> my pfsense does that I don t know why
[21:46] <alphe> but some time the network cards are only receiving data and can t send anything ...
[21:46] * dosaboy (~dosaboy@host109-155-13-93.range109-155.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[21:47] <alphe> i don t know if you will see that in the log that there is data send to the osd08 and that it never confirms reception
[21:48] <jhujhiti> alphe: all right, i'll play it safe then ;) just makes the upgrade a bit more painful later
[21:48] <alphe> jhujhiti or don t upgrade something that works great for you :)
[21:49] <jhujhiti> a memory leak in bobtail is biting us, so i will definitely be upgrading
[21:49] * sjm (~sjm@38.122.20.226) has joined #ceph
[21:49] <alphe> if it works if it does what you need in the parameters you need then don t upgrade
[21:49] <jhujhiti> i have to restart the OSDs every day
[21:49] <alphe> ok that is a good reason to upgrade
[21:50] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[21:59] <ircolle> jhujhiti - a point release for Dumpling should be out in next couple of days, if you can stand it that long, I'd wait until that comes out to upgrade
[21:59] <jhujhiti> it's not high priority for me now, more of an annoyance
[21:59] <jhujhiti> so, yes
[21:59] <jhujhiti> i can wait
[22:00] * portante is now known as portante|afk
[22:00] * portante|afk is now known as portante
[22:07] * sagelap (~sage@2607:f298:a:607:ea03:9aff:febc:4c23) has joined #ceph
[22:07] <sprachgenerator> alphe so I have run just a single mon create and now the ceph-create-keys process just appears hung: http://pastebin.com/gvfyAWd5 - been running for about 4 minutes now
[22:09] * allsystemsarego (~allsystem@188.27.167.103) Quit (Quit: Leaving)
[22:12] * sjm (~sjm@38.122.20.226) has joined #ceph
[22:13] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) has joined #ceph
[22:14] * gregaf1 (~Adium@cpe-172-250-66-79.socal.res.rr.com) Quit (Quit: Leaving.)
[22:16] <alphe> sprachgenerator that is normal
[22:16] <sagewk> zackc: we'll still need the helper (or a different helper) to enable coredumps
[22:16] <rsanti> rovar: but if I _do_ enumerate them in conf...
[22:16] <rsanti> sorry for the... slightly late reply
[22:16] <alphe> now create the other ones
[22:16] <sprachgenerator> I have - quorum is now established according to the logs
[22:16] <rsanti> (I have to generate the conf file anyway.)
[22:17] <alphe> sparchgenerator ok then the next step is to gatherkeys that allows you to test if all is ok
[22:17] <sprachgenerator> same issue - but the create keys process is still there
[22:18] <dmick> sprachgenerator: see what it's logging, with strace if nothing else
[22:19] <zackc> sagewk: maybe we can use it less often. or put it in $PATH. or drop the ceph-coverage stuff if we're not using it
[22:20] <zackc> we currently have ~100 characters of helper noise before many of the commands we execute via teuthology
[22:20] <sprachgenerator> here is the output of a manual key create: http://pastebin.com/yj92YUBF
[22:20] <zackc> and parsing is hard enough
[22:20] <sprachgenerator> so in this case
[22:21] <sprachgenerator> checking that connectivity, I see that even though the hostname is setup and listed properly, the other ceph nodes are listening on another interface
[22:21] <sprachgenerator> is there a way to force them to listen on all interfaces?
[22:21] <sprachgenerator> or a specific one?
[22:21] * grepory (~Adium@2600:1010:b026:2469:9194:f954:cf82:120) has joined #ceph
[22:21] * jhujhiti (~jhujhiti@00012a8b.user.oftc.net) has left #ceph
[22:22] * ScOut3R (~scout3r@BC2484D1.dsl.pool.telekom.hu) Quit (Remote host closed the connection)
[22:23] <alphe> is there a firewall running on your nodes sprachgenerator
[22:24] * danieagle (~Daniel@177.99.135.65) has joined #ceph
[22:24] <alphe> selinux or iptables or both
[22:25] <alphe> what key create says is that it can t connect to 172.21.0.109:6789
[22:25] <alphe> or 172.21.0.108:6789
[22:25] <alphe> or 172.21.0.110:6789
[22:26] <alphe> my guess is that or the monitors are not running or that those ports are firbiden by anytime of firewall
[22:26] <sagewk> yeah some of the other stuff would be nice to drop.
[22:26] <sagewk> we could try to wrap it all up into a single wrapper instead of chaining them together..
[22:28] <sagewk> zackc: ^
[22:28] <alphe> ircolle so far there was no freeze of the ceph-fuse so I assume it was my weird bug
[22:29] <zackc> sagewk: yeah that would be another option
[22:30] <ircolle> alphe - glad to hear it
[22:31] <alphe> but still that is an imprevisible bug for me
[22:31] <alphe> it is not ceph related but it is impacting ceph
[22:31] <sprachgenerator> alphe I was able to sort it out - thanks much for you help! on to the OSD's
[22:31] <alphe> i will have to see with the pfsense guys ...
[22:32] * gregaf1 (~Adium@cpe-172-250-66-79.socal.res.rr.com) has joined #ceph
[22:32] <alphe> sprachgenerator you are welcome :) comeback anytime you will get answers here
[22:32] <sprachgenerator> alphe it was really two parts - first the fact that I was trying to create all mons at once (as you stated) and the second was that the purge/purgedata needed to be run otherwise the mon nodes were still listening on the old interfaces
[22:34] <alphe> :)
[22:34] <alphe> glad you get it working :)
[22:37] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) Quit (Quit: tryggvil)
[22:39] <sagelap> dmick: https://github.com/ceph/ceph/pull/565
[22:39] * portante is now known as portante|afk
[22:40] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) has joined #ceph
[22:41] * sleinen (~Adium@2001:620:0:25:48d4:6aa9:7ce9:790a) has joined #ceph
[22:41] <dmick> sagewk: are all the arms fpu=neon, or should we be testing more specifically?...
[22:42] <sagewk> dmick: that was the magic that calxeda gave us
[22:45] * lx0 (~aoliva@lxo.user.oftc.net) Quit (Remote host closed the connection)
[22:45] * tobru_ (~quassel@217-162-50-53.dynamic.hispeed.ch) has joined #ceph
[22:47] <dmick> NEON is included in all Cortex-A8 devices but is optional in Cortex-A9 devices. <shrug> Maybe the gcc generated code has its own fallback?...
[22:50] * rturk is now known as rturk-away
[22:50] <sagewk> oh, there is runtime detection
[22:50] <sagewk> see arch/neon.c
[22:52] * tobru_ (~quassel@217-162-50-53.dynamic.hispeed.ch) Quit (Remote host closed the connection)
[22:53] * nhorman (~nhorman@hmsreliant.think-freely.org) Quit (Quit: Leaving)
[22:54] * mschiff_ (~mschiff@port-46124.pppoe.wtnet.de) Quit (Remote host closed the connection)
[22:54] <alphe> I have a question what are the page number that I should give to each pool of a rgw s3 ?
[22:55] <alphe> I imagine the most important is .rgw.bucket that one has to weight most of the ceph storage no ?
[22:56] * sjm (~sjm@38.122.20.226) Quit (Remote host closed the connection)
[22:59] * mattt (~mattt@lnx1.defunct.ca) has joined #ceph
[23:07] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) Quit (Quit: Ex-Chat)
[23:10] * aliguori (~anthony@cpe-70-112-153-179.austin.res.rr.com) has joined #ceph
[23:12] * sjm (~sjm@38.122.20.226) has joined #ceph
[23:14] * sprachgenerator (~sprachgen@130.202.135.214) Quit (Quit: sprachgenerator)
[23:15] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) Quit (Ping timeout: 480 seconds)
[23:15] * sjm (~sjm@38.122.20.226) Quit (Read error: Connection reset by peer)
[23:15] * sjm (~sjm@2607:f298:a:607:5165:a808:d1c7:3e6f) has joined #ceph
[23:17] * sprachgenerator (~sprachgen@130.202.135.214) has joined #ceph
[23:18] * sprachgenerator (~sprachgen@130.202.135.214) Quit ()
[23:19] * markbby (~Adium@168.94.245.1) Quit (Quit: Leaving.)
[23:21] <sagewk> zackc: https://github.com/ceph/teuthology/pull/71
[23:24] <zackc> looks good
[23:24] <alphe> I have a question what are the page number that I should give to each pool of a rgw s3 ?
[23:26] <dmick> sagewk: having trouble understanding the impact of the initial value, but it smells like it's arbitrary?...can we migrate to using the friendlier one? (I see other implementations of CRC32C that seem to imply that ~0 is common)
[23:26] <sagewk> dmick: i don't think we can change it now that previous versions are using 0
[23:27] <sagewk> and anyway, it doesn't really matter since we'll want ot compose multiple crcs together so it can have arbitrary values
[23:27] <sagewk> you only get to pick it at teh beginning of the buffer
[23:27] <dmick> k
[23:28] <dmick> happier with the fpu only for libcrc, as long as that's safe (there's state-save/restore code for XMM on x86, for instance)
[23:28] <sagewk> alphe: http://ceph.com/docs/master/rados/operations/placement-groups/
[23:28] <dmick> (I know, I just pointed it out, but I don't actually know the answer)
[23:28] <sagewk> alphe: make the rgw data pool big, the others can be smallish
[23:29] * vata (~vata@2607:fad8:4:6:b93b:c4fe:b6d0:d484) Quit (Quit: Leaving.)
[23:29] <alphe> ok but smallish to what extend 1 page ?
[23:29] <alphe> 2 page ?
[23:30] <dmick> pg == placement group
[23:30] <alphe> my ceph.conf generated by ceph-deploy don t show limitations recommanded in the osd pool management doc ...
[23:30] <alphe> dmick yes sorry :)
[23:30] * danieagle (~Daniel@177.99.135.65) Quit (Quit: inte+ e Obrigado Por tudo mesmo! :-D)
[23:31] <alphe> I m a bit confuse
[23:31] <alphe> placement group
[23:31] * dosaboy_ is now known as dosaboy
[23:31] * thomnico (~thomnico@2a01:e35:8b41:120:dc85:9784:3cef:7288) has joined #ceph
[23:40] * portante|afk is now known as portante
[23:41] * grepory (~Adium@2600:1010:b026:2469:9194:f954:cf82:120) Quit (Quit: Leaving.)
[23:44] * vanham (~vanham@187.115.200.154) has joined #ceph
[23:45] <vanham> Folks, does anyone knows how to remove MDSs from my cluster
[23:45] <vanham> I want to remove all the used data
[23:45] <vanham> I want to format my CephFS
[23:45] <vanham> Restart my MDSs
[23:45] <vanham> *Reinitialize my MDSs
[23:46] <sagewk> vanham: delete the pools, create a new data and metadata pool, figure out the numeric ids (ceph osd dump | grep ^pool), and use the 'ceph mds newfs ...' command
[23:46] <sagewk> (delete the data and metadata pools; other pools can be left behind.)
[23:47] <vanham> Thanks sage
[23:47] <vanham> Okay, I just have to move the RBD images away from the data pool first (my mistake here)
[23:47] <lurbs> sagewk: Odd question, but how do you pronounce your last name? I'm supposed to be doing a talk on Ceph in the near future, and would like to get it right.
[23:48] <sagewk> you can leave the old fs pools behidn if you want.
[23:48] <sagewk> maybe rename then (ceph osd pool rename ...)
[23:48] <sagewk> just to avoid confusion
[23:48] <dmsimard1> So, I know about pools. About replicas. About physical topology in the crush map (Datacenter -> Row -> Rack -> Host -> OSD). I'm trying to find in the documentation how to, for example, make sure replicas are at least made on another host. Any ideas ?
[23:49] * dmsimard1 is now known as dmsimard
[23:49] <vanham> sagewk, great idea! thanks!
[23:49] <alfredodeza> I am implementing a flag in ceph-deploy to avoid modifying package repositories so that you can only install ceph+dependencies from whatever configured repos already exist
[23:49] <alfredodeza> but I am having a hard time coming up with a decent CLI flag
[23:50] <sagewk> --adjust-package-sources, --no-adjust-package-sources ?
[23:50] <alfredodeza> currently I am using `--pkgs-only` so that the install command would look like: `ceph-deploy install --pkgs-only {host}`
[23:50] <sagewk> --adjust-sources / --no-adjust-sources ?
[23:50] <alfredodeza> hrmn
[23:50] <sagewk> dmsimard: ceph separates acrsos hosts by default
[23:50] <lurbs> dmsimard: That's also in the CRUSH map, as per: http://ceph.com/docs/master/rados/operations/crush-map/#crush-map-rules
[23:51] * imjustmatthew_ (~imjustmat@pool-173-53-100-217.rcmdva.fios.verizon.net) has joined #ceph
[23:51] <dmick> --use-current-pkgrepos
[23:52] <sagewk> --adjust-repo / --no-adjust-repo ? pkg is implied for an install command ..
[23:53] <sagewk> dmsimard: you can also set it on cluster creation via 'osd crush chooseleaf type = 1' (host) or 2 (rack) etc
[23:53] <sagewk> dmsimard: or you can create new crush rules with ceph osd crush rule .... or manually extract, modify, and reinject the crush map.
[23:55] <dmsimard> sagewk, lurbs, thanks - that sheds some light
[23:56] <dmsimard> I guess what I was trying to find is additional documentations on how rulesets work
[23:56] * BillK (~BillK-OFT@58-7-131-166.dyn.iinet.net.au) has joined #ceph
[23:56] * mtl1 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) has left #ceph
[23:57] * jeff-YF (~jeffyf@67.23.117.122) Quit (Quit: jeff-YF)
[23:58] <zackc> --preserve-repos ?
[23:58] * mattt (~mattt@lnx1.defunct.ca) Quit (Quit: leaving)
[23:59] <dmsimard> It looks like what I'm looking for is around step choose/step chooseleaf
[23:59] <sagewk> alfredodeza: i think anything that is in reference to setting or not setting up the repo sources is fine. i don't like --packages-only (or variations) because it's not obvious what that is(n't)
[23:59] <sagewk> dmsimard: the type argument to chooseleaf, right
[23:59] <alfredodeza> yep
[23:59] <alfredodeza> hrmn
[23:59] <alfredodeza> --preserve-repos sounds good
[23:59] * alfredodeza changes that
[23:59] <alfredodeza> norris zackc
[23:59] <kraken> When Alexander Bell invented the telephone he had 3 missed calls from zackc

These logs were automatically created by CephLogBot on irc.oftc.net using the Java IRC LogBot.