#ceph IRC Log

Index

IRC Log for 2013-08-20

Timestamps are in GMT/BST.

[0:00] <sagewk> yeah
[0:00] <Fetch_> I ask because I applied it manually (just added the 1 line to the ceph command) and I get AttributeError: 'Rados' object has no attribute 'conf_parse_env'
[0:00] <sagewk> you also need the previous patch that adds the call to rados.py
[0:01] <Fetch_> gotcha, I see it
[0:01] <Fetch_> that's what I get for going off the bugrep instead of the github
[0:04] <sagewk> :)
[0:04] <sagewk> sjust: wip-osd-mode when you have a few minutes to talk
[0:06] <dmick> it's brand spankin' new Fetch_
[0:07] <Fetch_> dmick: the patch? yes, very. But a long as I'm debugging OS/Ceph/quantum issues, I might as well get that one out of the way (because it can be pretty big for Cinder)
[0:07] <Fetch_> as long* I swear I'm typing like an idiot today
[0:07] <dmick> sure. I meant the conf_parse_env in rados.py
[0:21] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) Quit (Quit: Ja odoh a vi sta 'ocete...)
[0:24] * BillK (~BillK-OFT@58-7-52-33.dyn.iinet.net.au) has joined #ceph
[0:24] * roald (~oftc-webi@87.209.150.214) Quit (Quit: Page closed)
[0:26] * ShaunR- (~ShaunR@staff.ndchost.com) has joined #ceph
[0:28] * ShaunR (~ShaunR@staff.ndchost.com) Quit (Ping timeout: 480 seconds)
[0:31] * mschiff (~mschiff@85.182.236.82) Quit (Remote host closed the connection)
[0:38] * tnt (~tnt@109.130.104.131) Quit (Ping timeout: 480 seconds)
[0:40] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) Quit (Quit: Leaving...)
[0:42] * eternaleye (~eternaley@2002:3284:29cb::1) Quit (Read error: Connection reset by peer)
[0:43] * eternaleye (~eternaley@2002:3284:29cb::1) has joined #ceph
[0:43] * aliguori (~anthony@cpe-70-112-157-87.austin.res.rr.com) Quit (Remote host closed the connection)
[0:44] * AfC (~andrew@2407:7800:200:1011:f5ab:1454:3a67:ea29) has joined #ceph
[0:50] * Machske (~Bram@81.82.216.124) Quit (Read error: Connection reset by peer)
[0:50] * Machske (~Bram@d5152D87C.static.telenet.be) has joined #ceph
[0:53] * dalegaar1 (~dalegaard@vps.devrandom.dk) has joined #ceph
[0:53] * dalegaard (~dalegaard@vps.devrandom.dk) Quit (Read error: Connection reset by peer)
[0:53] * jksM (~jks@3e6b5724.rev.stofanet.dk) has joined #ceph
[0:53] * jks (~jks@3e6b5724.rev.stofanet.dk) Quit (Read error: Connection reset by peer)
[0:54] * Ilya_Bolotin (~ibolotin@38.122.20.226) Quit (Quit: Leaving.)
[0:54] <Psi-Jack> Hmm, 0.61.8 is out and /topic says 0.61.7?
[0:54] * Ilya_Bolotin (~ibolotin@38.122.20.226) has joined #ceph
[0:55] * alram (~alram@38.122.20.226) has joined #ceph
[0:55] * ChanServ sets mode +o sagewk
[0:55] * sagewk changes topic to 'Latest stable (v0.67.1 "Dumpling" or v0.61.8 "Cuttlefish") -- http://ceph.com/get || CDS Vids and IRC logs posted http://ceph.com/cds/'
[0:55] <Psi-Jack> hehe
[0:56] <Psi-Jack> Dumpling? Cuttlefish?
[0:56] <dmick> those are two major release names
[0:56] <dmick> Dumpling is the new hotness
[0:57] <Psi-Jack> yeah, but those are the same major version. :)
[0:57] <Psi-Jack> Ohhh waiyt.
[0:57] <Psi-Jack> I'm blind. :)
[0:57] <Psi-Jack> Hmmm... 0.67.1 already, eh? Hmmmm... Another named version.
[0:58] <dmick> Psi-Jack: we had several rc's before calling it released
[0:59] <Psi-Jack> Interesting issue I found and seemed to have so far resolved with an update from 0.61.6->0.61.8, I had a cephfs mount one one root for /home, and then another cephfs mount of a different root on /home/public, and somehow it lost public.
[1:00] <Psi-Jack> dmick: Oh yeah. I know. It's cool. I'll be getting up on it likely as soon as Proxmox VE catches up. :)
[1:00] <Psi-Jack> Since Proxmox devs seem to be on top of ceph these days.
[1:04] <Psi-Jack> Well, either way, regarding the issue I found, but got solevd with an update, I'll keep an eye out on if it happens again.
[1:05] <Psi-Jack> Was wierd, very wierd.. That a submount on to of a cephfs mount acted up and completely lost the directory, however, not the data, just the mount point itself.
[1:06] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) Quit (Ping timeout: 480 seconds)
[1:09] <jmlowe> I'm thinking we need a quick 0.67.2 release just to change the number for dyslexia avoidance
[1:10] <sagewk> mikedawson: ping
[1:10] <sagewk> jmlowe: seriously
[1:11] <jmlowe> I'd be surprised if mikedawson is around, it's 7:12 here
[1:12] <jmlowe> sagewk: what's the next bin up from 128TB for a support contract?
[1:12] <sagewk> probably 256? no idea honestly :)
[1:12] <sagewk> fetch_: how do you tell where an instaslled rpm came from with yum?
[1:13] <dmick> yum info <pkg> is verbose but works
[1:13] * gregmark (~Adium@68.87.42.115) Quit (Quit: Leaving.)
[1:13] <jmlowe> :) that's why I think you guys will do well, tech first money second
[1:15] <cmdrk> gregaf , sagewk : regarding my problem earlier, if there's a regression that is causing the issue, what kernel should i roll back to in the meanwhile? (i'm working on getting the ceph-client.git kernel built but i'd like to continue testing with a previous stable kernel if possible)
[1:16] <cmdrk> i have a 3.7 kernel available that i can easily roll back to at the moment
[1:17] <sagewk> you could try going back to 3.9 or 3.8 but it's a bit of a guess at this point. lots of fixes are trickling in to the fs stuff but we haven't been backporting them to stable kernels, so it's unclear what is broken/fixed when. aside from the dprune patch that hopefully addresses this, nothing is ringing bells
[1:17] <sagewk> glowell: hmm, seeing the same mon crash on centos63 with leveldb from centos6-qemu-local :/
[1:18] <cmdrk> thanks
[1:24] <glowell> is centos6-qemu-local the teuthology testing repo ?
[1:27] <glowell> Looks like it has leveldb-1.12. Updated on 16 August.
[1:34] <mikedawson> sagewk: I'm here now
[1:34] <sagewk> mikedawson: hey! just wondering if you've had any luck hitting that qemu hang
[1:34] <sagewk> or if the temporary hang is still the best we have to go on right now
[1:35] <mikedawson> sagewk: I gave joshd a log of one of the mystery dips, but I am 100% hang-free since the upgrade to qemu 1.5.2
[1:36] <mikedawson> sagewk: in my mind, issue #5955 is solved by the async flush patch in qemu
[1:40] <sagewk> hmm, ok. i haven't followed the qemu side but will check with josh when he's in tomorrow. i *did* still see something weird in the log from before :/
[1:40] <sagewk> there is a long-standing hang we were seeing with iozone, so maybe that is where I should go bug hunting
[1:42] <Fetch_> sagewk: dmick is right, also yum list <pkg> or yum list --showduplicates <pkg> if you want to see everything available
[1:43] <sagewk> glowell: teuthology testing repo?
[1:43] <sagewk> oh, so that's also our package.
[1:43] * mozg (~andrei@host109-151-35-94.range109-151.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[1:43] <glowell> Yes, it's a mirror that Sandon set up for teuthology.
[1:44] <sagewk> mirror of upstream centos, or our packages?
[1:44] <mikedawson> sagewk: I feel like we've seen most of the hangs/hiccups that have been reported due to our constant small random write workload.
[1:45] <sagewk> glowell: btw rhel6.4 also fails but it hangs instead of asserting
[1:47] <glowell> They are built on rhel6.3 so potentially the libraries liked to could be a different patch levels.
[1:53] <Fetch_> do you have the SRPM or a specfile somewhere I could download it?
[2:00] <glowell> For leveldb ? It's just a back port of the fedora19 package. You can find our copy at http://gitbuilder.ceph.com/packages/sources/leveldb/leveldb-1.12.0-3.fc19.src.rpm
[2:00] <glowell> The spec file was modified to add a build root and remove the VCS tag so it would build on older platfroms.
[2:03] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) Quit (Quit: Leaving.)
[2:05] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) has joined #ceph
[2:06] * alram (~alram@38.122.20.226) Quit (Ping timeout: 480 seconds)
[2:16] * sagelap (~sage@2607:f298:a:607:7067:94d7:68ab:4e08) Quit (Ping timeout: 480 seconds)
[2:20] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[2:21] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[2:23] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) Quit (Quit: Leaving...)
[2:24] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) has joined #ceph
[2:35] <Fetch_> glowell: I built that SRPM on a centos 6.4 box, installed to another 6.4 box running ceph mon, same error
[2:36] <Fetch_> of note, make check (done by rpmbuild by default) failed one of the checks, I couldn't tell if it was something likely to indicate the culprit
[2:36] <glowell> Interesting. I'll start checking for known bugs in the current version. Thnaks.
[2:37] <glowell> I don't remember seeing a problem with make check, but I turned it off for the production build.
[2:38] <Fetch_> I think it was one of the corruption tests
[2:39] <Fetch_> and the leveldb issue DB says it's a known race condition in tests, so no biggy
[2:47] * Cube1 (~Cube@88.128.80.12) has joined #ceph
[2:47] * Cube (~Cube@88.128.80.12) Quit (Read error: Connection reset by peer)
[2:47] * yanzheng (~zhyan@134.134.139.70) has joined #ceph
[2:50] * alram (~alram@cpe-76-167-50-51.socal.res.rr.com) has joined #ceph
[2:54] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) Quit (Quit: Leaving...)
[2:55] * yy-nm (~Thunderbi@122.233.231.235) has joined #ceph
[2:56] * haomaiwang (~haomaiwan@notes4.com) has joined #ceph
[2:57] * haomaiw__ (~haomaiwan@li565-182.members.linode.com) has joined #ceph
[3:01] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Ping timeout: 480 seconds)
[3:03] * haomaiwa_ (~haomaiwan@notes4.com) Quit (Ping timeout: 480 seconds)
[3:04] * haomaiwang (~haomaiwan@notes4.com) Quit (Ping timeout: 480 seconds)
[3:10] * huangjun (~kvirc@221.234.156.240) has joined #ceph
[3:12] <huangjun> Dumpling and Cuttlefish, which is the newest version?
[3:13] <sjust> Dumpling
[3:13] <Fetch_> dumpling
[3:13] <Fetch_> alphabetical order
[3:13] <Fetch_> although cuttlefish has a slightly newer point release
[3:36] * smiley (~smiley@cpe-67-251-108-92.stny.res.rr.com) has joined #ceph
[3:39] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[3:54] * clayb (~kvirc@69.191.241.59) Quit (Read error: Connection reset by peer)
[3:57] * dpippenger (~riven@tenant.pas.idealab.com) Quit (Remote host closed the connection)
[4:04] * grepory (~Adium@50-115-70-146.static-ip.telepacific.net) Quit (Quit: Leaving.)
[4:24] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Remote host closed the connection)
[4:26] * alram (~alram@cpe-76-167-50-51.socal.res.rr.com) Quit (Ping timeout: 480 seconds)
[4:32] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Ping timeout: 480 seconds)
[4:32] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) has joined #ceph
[4:42] * Cube (~Cube@88.128.80.12) has joined #ceph
[4:42] * Cube1 (~Cube@88.128.80.12) Quit (Read error: Connection reset by peer)
[4:42] * madkiss (~madkiss@184.105.243.169) Quit (Quit: Leaving.)
[4:51] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[4:52] * morse (~morse@supercomputing.univpm.it) Quit (Remote host closed the connection)
[4:56] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) has joined #ceph
[5:05] * fireD (~fireD@93-139-168-118.adsl.net.t-com.hr) has joined #ceph
[5:07] * fireD_ (~fireD@93-139-190-39.adsl.net.t-com.hr) Quit (Ping timeout: 480 seconds)
[5:07] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[5:11] * KindTwo (~KindOne@50.96.226.67) has joined #ceph
[5:13] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[5:13] * KindTwo is now known as KindOne
[5:14] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) has joined #ceph
[5:16] * jlogan (~Thunderbi@2600:c00:3010:1:1::40) has joined #ceph
[5:22] * jlogan1 (~Thunderbi@2600:c00:3010:1:1::40) Quit (Ping timeout: 480 seconds)
[5:22] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) has joined #ceph
[5:34] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) Quit (Quit: Leaving.)
[5:40] * haomaiwang (~haomaiwan@notes4.com) has joined #ceph
[5:42] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) has joined #ceph
[5:44] * haomaiw__ (~haomaiwan@li565-182.members.linode.com) Quit (Ping timeout: 480 seconds)
[5:53] * yy-nm (~Thunderbi@122.233.231.235) Quit (Quit: yy-nm)
[5:55] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) has joined #ceph
[5:55] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[6:02] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) has joined #ceph
[6:08] * rongze (~quassel@117.79.232.249) has joined #ceph
[6:12] * LPG (~LPG@c-76-104-197-224.hsd1.wa.comcast.net) Quit (Remote host closed the connection)
[6:13] * bandrus (~Adium@cpe-76-95-217-129.socal.res.rr.com) Quit (Quit: Leaving.)
[6:30] * rongze_ (~quassel@li565-182.members.linode.com) has joined #ceph
[6:34] * haomaiwa_ (~haomaiwan@notes4.com) has joined #ceph
[6:36] * rongze (~quassel@117.79.232.249) Quit (Ping timeout: 480 seconds)
[6:39] * haomaiwang (~haomaiwan@notes4.com) Quit (Ping timeout: 480 seconds)
[6:40] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) has joined #ceph
[6:48] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) Quit (Ping timeout: 480 seconds)
[7:00] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[7:01] * yy-nm (~Thunderbi@122.233.231.235) has joined #ceph
[7:06] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[7:08] * buck (~buck@c-24-6-91-4.hsd1.ca.comcast.net) has left #ceph
[7:18] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) has joined #ceph
[7:21] * sleinen1 (~Adium@2001:620:0:26:a0f4:21b5:eb88:bb82) has joined #ceph
[7:26] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[7:27] * sleinen (~Adium@77-58-245-10.dclient.hispeed.ch) Quit (Ping timeout: 480 seconds)
[7:30] * sleinen (~Adium@2001:620:0:26:7d34:6437:2d21:7413) has joined #ceph
[7:33] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) Quit (Quit: Leaving.)
[7:34] * KindOne (~KindOne@0001a7db.user.oftc.net) has joined #ceph
[7:37] * sleinen1 (~Adium@2001:620:0:26:a0f4:21b5:eb88:bb82) Quit (Ping timeout: 480 seconds)
[7:37] * haomaiwang (~haomaiwan@li565-182.members.linode.com) has joined #ceph
[7:41] * haomaiwa_ (~haomaiwan@notes4.com) Quit (Ping timeout: 480 seconds)
[7:46] * Machske (~Bram@d5152D87C.static.telenet.be) Quit ()
[7:46] * huangjun|2 (~kvirc@221.234.156.240) has joined #ceph
[7:47] * torment1 (~torment@pool-72-64-180-146.tampfl.fios.verizon.net) Quit (Read error: Operation timed out)
[7:47] * sleinen (~Adium@2001:620:0:26:7d34:6437:2d21:7413) Quit (Quit: Leaving.)
[7:48] * huangjun (~kvirc@221.234.156.240) Quit (Ping timeout: 480 seconds)
[7:55] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) has joined #ceph
[7:58] * mtl2 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) Quit (Read error: Operation timed out)
[8:01] * xmltok (~xmltok@cpe-76-170-26-114.socal.res.rr.com) Quit (Quit: Bye!)
[8:06] * haomaiwang (~haomaiwan@li565-182.members.linode.com) Quit (Remote host closed the connection)
[8:06] * madkiss (~madkiss@184.105.243.169) Quit (Quit: Leaving.)
[8:10] * AfC1 (~andrew@2407:7800:200:1011:f5ab:1454:3a67:ea29) has joined #ceph
[8:10] * AfC (~andrew@2407:7800:200:1011:f5ab:1454:3a67:ea29) Quit (Read error: No route to host)
[8:13] * grepory (~Adium@c-69-181-42-170.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[8:18] * allsystemsarego (~allsystem@5-12-241-157.residential.rdsnet.ro) has joined #ceph
[8:25] * sleinen (~Adium@2001:620:0:30:9411:4385:b807:62d0) has joined #ceph
[8:27] * tnt (~tnt@109.130.104.131) has joined #ceph
[8:31] <huangjun|2> i have 3 mons,{a,b,c},and i stop a, mon cluster is ok bc the 'ceph -w' works fine, but if i stop b then, the 'ceph -w' shows nothing,
[8:33] * sleinen (~Adium@2001:620:0:30:9411:4385:b807:62d0) Quit (Ping timeout: 480 seconds)
[8:33] <huangjun|2> i think it is abnormal
[8:37] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[8:38] * sleinen (~Adium@2001:620:0:26:70b3:1035:6ee:1832) has joined #ceph
[8:39] * lxo (~aoliva@lxo.user.oftc.net) Quit (Remote host closed the connection)
[8:39] * sleinen (~Adium@2001:620:0:26:70b3:1035:6ee:1832) Quit ()
[8:39] * sleinen (~Adium@130.59.92.67) has joined #ceph
[8:44] <yanzheng> it's expected behavior
[8:45] <yanzheng> the only mon don't know if the two stopped mon have new information
[8:46] <yanzheng> so it can't give you any information
[8:47] * sleinen (~Adium@130.59.92.67) Quit (Ping timeout: 480 seconds)
[8:50] * madkiss (~madkiss@184.105.243.169) Quit (Ping timeout: 480 seconds)
[8:52] * Cube (~Cube@88.128.80.12) Quit (Quit: Leaving.)
[8:53] * lxo (~aoliva@lxo.user.oftc.net) has joined #ceph
[8:59] * sleinen (~Adium@130.59.92.67) has joined #ceph
[8:59] * sleinen (~Adium@130.59.92.67) Quit ()
[9:02] * sleinen (~Adium@130.59.92.67) has joined #ceph
[9:03] * hybrid512 (~walid@LPoitiers-156-86-25-85.w193-248.abo.wanadoo.fr) has joined #ceph
[9:03] * sleinen1 (~Adium@2001:620:0:26:58dd:e246:9723:948a) has joined #ceph
[9:06] * BManojlovic (~steki@91.195.39.5) has joined #ceph
[9:08] * indego (~indego@91.232.88.10) has joined #ceph
[9:10] * sleinen (~Adium@130.59.92.67) Quit (Ping timeout: 480 seconds)
[9:11] * tziOm (~bjornar@194.19.106.242) has joined #ceph
[9:14] * iggy (~iggy@theiggy.com) Quit (Quit: No Ping reply in 180 seconds.)
[9:14] * iggy (~iggy@theiggy.com) has joined #ceph
[9:19] * tnt (~tnt@109.130.104.131) Quit (Ping timeout: 480 seconds)
[9:20] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[9:20] * ChanServ sets mode +v andreask
[9:20] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[9:21] <yy-nm> yanzheng: do you know the value of macro ENCODE_DUMP which show in include/encoding.h L131
[9:24] <yanzheng> grep show it's defined in do_autogen.sh
[9:25] <yy-nm> yanzheng: thanks, i forget the file out src/
[9:25] * jlogan (~Thunderbi@2600:c00:3010:1:1::40) Quit (Ping timeout: 480 seconds)
[9:29] * madkiss (~madkiss@184.105.243.169) Quit (Ping timeout: 480 seconds)
[9:30] <huangjun|2> yanzheng: why? the only mon will have the newest information when there only are two mons, after another mon down, the only mon should be as leader and offering service
[9:31] * tnt (~tnt@212-166-48-236.win.be) has joined #ceph
[9:32] * JM (~oftc-webi@193.252.138.241) has joined #ceph
[9:32] * mnash (~chatzilla@vpn.expressionanalysis.com) Quit (Remote host closed the connection)
[9:32] * AfC1 (~andrew@2407:7800:200:1011:f5ab:1454:3a67:ea29) Quit (Quit: Leaving.)
[9:35] <yanzheng> huangjun|2, I think the mon implementation does not take the down sequence into consideration
[9:35] <yy-nm> not, monitor quorum must have more than half node alive, then you can get newest information
[9:35] <yanzheng> It just checks how many mon are active
[9:36] * Cube (~Cube@et-0-30.gw-nat.bs.kae.de.oneandone.net) has joined #ceph
[9:38] * BManojlovic (~steki@91.195.39.5) Quit (Quit: Ja odoh a vi sta 'ocete...)
[9:38] * BManojlovic (~steki@91.195.39.5) has joined #ceph
[9:41] <yy-nm> damn it, ENCODE_DUMP seems not exist in 0.61.4 package.
[9:42] <yanzheng> why do you need it
[9:45] <yy-nm> i read the source code to understand ceph behavior deeply.
[9:47] * odyssey4me3 (~odyssey4m@165.233.71.2) has joined #ceph
[9:49] * jlogan (~Thunderbi@2600:c00:3010:1:1::40) has joined #ceph
[9:50] * BManojlovic (~steki@91.195.39.5) Quit (Quit: Ja odoh a vi sta 'ocete...)
[9:53] * odyssey4me3 is now known as odyssey4me
[9:55] * BManojlovic (~steki@91.195.39.5) has joined #ceph
[10:04] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[10:11] * madkiss (~madkiss@184.105.243.169) Quit (Read error: Operation timed out)
[10:12] * tnt_ (~tnt@212-166-48-236.win.be) has joined #ceph
[10:14] * tnt (~tnt@212-166-48-236.win.be) Quit (Ping timeout: 480 seconds)
[10:24] * KindTwo (~KindOne@h150.173.17.98.dynamic.ip.windstream.net) has joined #ceph
[10:25] * sleinen1 (~Adium@2001:620:0:26:58dd:e246:9723:948a) Quit (Quit: Leaving.)
[10:26] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[10:26] * KindTwo is now known as KindOne
[10:30] * jbd_ (~jbd_@2001:41d0:52:a00::77) has joined #ceph
[10:39] <loicd> morning ceph (evening yanzheng ;-)
[10:39] <yanzheng> morning loicd
[10:43] <ccourtaut> mornin
[10:46] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[10:47] * KindOne (~KindOne@0001a7db.user.oftc.net) has joined #ceph
[10:48] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[10:52] <joelio> mornin'!
[10:54] * madkiss (~madkiss@184.105.243.169) Quit (Read error: Operation timed out)
[10:56] * yanzheng (~zhyan@134.134.139.70) Quit (Remote host closed the connection)
[11:03] * sleinen (~Adium@user-28-18.vpn.switch.ch) has joined #ceph
[11:03] * julian (~julianwa@125.69.106.188) has joined #ceph
[11:15] * appelza (~pieter@105-236-207-237.access.mtnbusiness.co.za) has joined #ceph
[11:17] <appelza> Is it feasible/wise to use cephfs for a simple network raid1 type of design? 2x machines, 2x block devices replicated via ceph, can handle failover of either machine?
[11:17] <appelza> using cephfs probably
[11:18] <appelza> (I know it's possible.)
[11:23] <loicd> appelza: I plan to do exactly that (but with RBD) using two servers in my basement, as a replacement for DRBD.
[11:23] <loicd> for backups
[11:27] * vanham (~vanham@209.156.172.51) has joined #ceph
[11:27] <vanham> Morning everyone
[11:27] <vanham> I have a production cluster running cuttlefish (0.61.7) but my MDS is stuck at replay. It is kind of urgent
[11:27] <vanham> Does anyone have any suggestion?
[11:30] <appelza> cool :)
[11:30] <vanham> it is "waiting for osdmap"
[11:31] <loicd> vanham: I won't be able to help you much but when someone more knowledgeable shows up, it will be useful to have logs + your configuration stored somewhere for analysis.
[11:32] <vanham> Config: http://pastebin.com/n8p3yhVX
[11:32] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[11:32] <vanham> MDS log: http://pastebin.com/SxD03hYH
[11:33] * gionn (~oftc-webi@host40-30-static.224-95-b.business.telecomitalia.it) has joined #ceph
[11:34] <gionn> hi, deploying ceph via chef is still supported?
[11:35] * rongze (~quassel@117.79.232.217) has joined #ceph
[11:36] <vanham> my MDSs are getting blacklisted
[11:36] <vanham> Why is it?
[11:36] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) Quit (Ping timeout: 480 seconds)
[11:38] <vanham> loicd, any idea?
[11:39] * mtl1 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) has joined #ceph
[11:39] <loicd> vanham: I won't be much of a help I'm afraid, this is new to me :-) What does ceph -s say ?
[11:40] * madkiss (~madkiss@184.105.243.169) Quit (Ping timeout: 480 seconds)
[11:40] <loicd> gionn: I don't use it myself but I know it is supported :-)
[11:41] <vanham> loicd,
[11:41] <vanham> http://pastebin.com/3PDp7Qk4
[11:41] <vanham> loicd, http://pastebin.com/3PDp7Qk4
[11:42] <vanham> Is there anyone I could call?
[11:43] * rongze_ (~quassel@li565-182.members.linode.com) Quit (Ping timeout: 480 seconds)
[11:44] <loicd> vanham: do you have a support contract with someone ?
[11:44] <vanham> No
[11:44] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[11:45] * KindOne (~KindOne@0001a7db.user.oftc.net) has joined #ceph
[11:46] <loicd> if it's critical and you're ready to spend some money to get this fixed asap, I can think of two europe based companies who provide that kind of service : http://www.hastexo.com/ and https://www.42on.com/
[11:47] <loicd> ( I'm not affiliated to them ;-)
[11:48] <vanham> thanks
[11:52] * yy-nm (~Thunderbi@122.233.231.235) Quit (Remote host closed the connection)
[11:54] * mtl2 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) has joined #ceph
[11:55] * vanham (~vanham@209.156.172.51) Quit (Remote host closed the connection)
[11:59] * odyssey4me (~odyssey4m@165.233.71.2) Quit (Ping timeout: 480 seconds)
[11:59] * mtl1 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) Quit (Ping timeout: 480 seconds)
[11:59] * Cube (~Cube@et-0-30.gw-nat.bs.kae.de.oneandone.net) Quit (Quit: Leaving.)
[12:04] * odyssey4me (~odyssey4m@165.233.205.190) has joined #ceph
[12:06] * sleinen1 (~Adium@eduroam-hg-dock-1-37.ethz.ch) has joined #ceph
[12:09] * sleinen2 (~Adium@2001:620:0:25:14e9:9db0:7655:1795) has joined #ceph
[12:09] * tserong_ (~tserong@124-168-231-241.dyn.iinet.net.au) has joined #ceph
[12:13] * lightspeed (~lightspee@81.187.0.153) Quit (Ping timeout: 480 seconds)
[12:13] * sleinen (~Adium@user-28-18.vpn.switch.ch) Quit (Ping timeout: 480 seconds)
[12:15] * sleinen1 (~Adium@eduroam-hg-dock-1-37.ethz.ch) Quit (Ping timeout: 480 seconds)
[12:15] * mozg (~andrei@host217-46-236-49.in-addr.btopenworld.com) has joined #ceph
[12:16] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[12:16] * tserong (~tserong@124-171-113-175.dyn.iinet.net.au) Quit (Ping timeout: 480 seconds)
[12:17] <gionn> loicd: I'm trying to deploy an entire ceph cluster on a single node for testing purposes with chef, but it get stuck after starting the first mon, when it try to execute 'ceph auth get-key client.bootstrap-osd'
[12:17] <gionn> I've read on the documentation the notice about the osd crush chooseleaf type = 0 (should be set on /etc/ceph/ceph.conf, right?)
[12:19] <gionn> any hints?
[12:20] * X3NQ (~X3NQ@195.191.107.205) has joined #ceph
[12:20] * odyssey4me (~odyssey4m@165.233.205.190) Quit (Ping timeout: 480 seconds)
[12:22] * KindTwo (~KindOne@198.14.201.128) has joined #ceph
[12:23] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[12:23] * KindTwo is now known as KindOne
[12:24] * madkiss (~madkiss@184.105.243.169) Quit (Ping timeout: 480 seconds)
[12:24] * odyssey4me (~odyssey4m@165.233.71.2) has joined #ceph
[12:25] * mtl1 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) has joined #ceph
[12:30] * mtl2 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) Quit (Ping timeout: 480 seconds)
[12:31] * vanham (~vanham@66.250.218.250) has joined #ceph
[12:45] * LeaChim (~LeaChim@176.24.168.228) has joined #ceph
[12:47] * sleinen2 (~Adium@2001:620:0:25:14e9:9db0:7655:1795) Quit (Quit: Leaving.)
[12:47] * sleinen (~Adium@eduroam-hg-dock-1-37.ethz.ch) has joined #ceph
[12:55] * Cube (~Cube@et-0-30.gw-nat.bs.kae.de.oneandone.net) has joined #ceph
[12:55] * sleinen (~Adium@eduroam-hg-dock-1-37.ethz.ch) Quit (Ping timeout: 480 seconds)
[12:57] * dosaboy_ (~dosaboy@host109-158-236-83.range109-158.btcentralplus.com) has joined #ceph
[12:58] <mozg> gionn, not really sure how to do it with chef, but I would initially start by creating a cluster manually to make sure it works, followed by the chef integration
[12:59] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[13:00] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) has joined #ceph
[13:03] * dosaboy (~dosaboy@host109-154-149-172.range109-154.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[13:05] * madkiss (~madkiss@184.105.243.169) Quit (Read error: Operation timed out)
[13:09] * huangjun|2 (~kvirc@221.234.156.240) Quit (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/)
[13:10] * yanzheng (~zhyan@101.83.110.113) has joined #ceph
[13:16] * mtl2 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) has joined #ceph
[13:16] * mtl1 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) Quit (Read error: Operation timed out)
[13:21] * vanham (~vanham@66.250.218.250) Quit (Remote host closed the connection)
[13:29] * lx0 (~aoliva@lxo.user.oftc.net) has joined #ceph
[13:30] * lxo (~aoliva@lxo.user.oftc.net) Quit (Remote host closed the connection)
[13:30] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[13:34] <loicd> you could start the cluster with a single mon gionn : that will make things even simpler for a test
[13:35] <loicd> and solve your problem at the same time, probably ;-)
[13:36] * sleinen (~Adium@130.59.92.68) has joined #ceph
[13:37] * sleinen1 (~Adium@2001:620:0:25:d91d:df90:c9fa:177f) has joined #ceph
[13:38] * madkiss (~madkiss@184.105.243.169) Quit (Ping timeout: 480 seconds)
[13:40] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) has joined #ceph
[13:44] * sleinen (~Adium@130.59.92.68) Quit (Ping timeout: 480 seconds)
[13:44] * sleinen1 (~Adium@2001:620:0:25:d91d:df90:c9fa:177f) Quit (Quit: Leaving.)
[13:54] * nhorman (~nhorman@hmsreliant.think-freely.org) has joined #ceph
[14:05] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[14:06] * jlogan2 (~Thunderbi@2600:c00:3010:1:1::40) has joined #ceph
[14:06] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[14:06] * jlogan (~Thunderbi@2600:c00:3010:1:1::40) Quit (Ping timeout: 480 seconds)
[14:09] * baffle_ (baffle@jump.stenstad.net) Quit (Ping timeout: 480 seconds)
[14:12] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) Quit (Ping timeout: 480 seconds)
[14:13] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[14:13] * baffle (baffle@jump.stenstad.net) has joined #ceph
[14:16] * mtl2 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) Quit (Ping timeout: 480 seconds)
[14:24] * markbby (~Adium@168.94.245.4) has joined #ceph
[14:26] * mtl1 (~Adium@c-67-176-54-246.hsd1.co.comcast.net) has joined #ceph
[14:31] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) has joined #ceph
[14:39] * joao (~joao@89.181.146.94) has joined #ceph
[14:39] * ChanServ sets mode +o joao
[14:39] * JM (~oftc-webi@193.252.138.241) Quit (Quit: Page closed)
[14:42] * pieter_ (~pieter@105-236-133-114.access.mtnbusiness.co.za) has joined #ceph
[14:46] * appelza (~pieter@105-236-207-237.access.mtnbusiness.co.za) Quit (Ping timeout: 480 seconds)
[14:52] * jefferai (~quassel@corkblock.jefferai.org) Quit (Ping timeout: 480 seconds)
[14:53] * pieter_ (~pieter@105-236-133-114.access.mtnbusiness.co.za) Quit (Quit: Konversation terminated!)
[14:55] * jefferai (~quassel@corkblock.jefferai.org) has joined #ceph
[14:58] * clayb (~kvirc@proxy-nj1.bloomberg.com) has joined #ceph
[15:03] * `10` (~10@juke.fm) has joined #ceph
[15:09] * jeff-YF (~jeffyf@67.23.117.122) has joined #ceph
[15:10] * `10 (~10@juke.fm) Quit (Ping timeout: 480 seconds)
[15:17] * Vjarjadian (~IceChat77@90.214.208.5) Quit (Quit: Depression is merely anger without enthusiasm)
[15:37] * torment1 (~torment@pool-96-228-147-37.tampfl.fios.verizon.net) has joined #ceph
[15:40] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) has joined #ceph
[15:44] * foosinn (~stefan@office.unitedcolo.de) has joined #ceph
[15:46] * yanzheng (~zhyan@101.83.110.113) Quit (Ping timeout: 480 seconds)
[15:46] <foosinn> hi together, i have setup a ceph cluster with 2 storage nodes, each one has a 9gb osd. as far as i understood ceph it takes automatically care for redundance. after mounting a cephfs i have a 18gb fs. am i wrong?
[15:47] * yanzheng (~zhyan@101.83.43.148) has joined #ceph
[15:48] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) Quit (Read error: Operation timed out)
[15:48] <tnt_> foosinn: no, it's normal.
[15:49] <tnt_> foosinn: ceph can't predict acurately free space because you can configure some zone to be replicated 2x,3x,4x,...
[15:49] <foosinn> tnt_, thanks for you help, do i have redundancy?
[15:49] <tnt_> try creating a 1Go file and you should see the free space be reduced by 2Go.
[15:52] * vipr (~vipr@78-23-114-68.access.telenet.be) Quit (Remote host closed the connection)
[15:52] * gionn (~oftc-webi@host40-30-static.224-95-b.business.telecomitalia.it) Quit (Quit: Page closed)
[15:54] <foosinn> ah ok, i unterstand. another question: can i only setup a single cephfs? i didnt mention a name or anything else, i just mounted it.
[15:54] * madkiss (~madkiss@184.105.243.169) Quit (Read error: Connection reset by peer)
[15:54] * madkiss (~madkiss@184.105.243.169) has joined #ceph
[15:55] <tnt_> yes you can only have one cephfs per cluster.
[15:55] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) has joined #ceph
[15:55] <tnt_> (atm, afaik)
[15:55] <tnt_> you can mount subdirectories though, you're not forced to mount the root.
[15:57] <foosinn> ok, one last question: is there a blog for the status of cephfs (i found a youtube vid where someone said 3 moths ago cephfs is close to stable)
[15:57] <foosinn> tnt_, i already tryed out that :) but thanks
[15:59] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[16:01] * KindOne (~KindOne@0001a7db.user.oftc.net) has joined #ceph
[16:01] <tnt_> same blog as ceph on ceph.com
[16:02] <foosinn> ok, seem like i'm just bild. thanks a lot for your help
[16:03] * aliguori (~anthony@32.97.110.51) has joined #ceph
[16:04] <ircolle> foosinn - "close to stable" is clearly in the eye of the beholder
[16:04] <joelio> stable for me, depends on your use case I guess
[16:05] <nhm> joelio: have you noticed any performance hicups?
[16:06] <joelio> nhm: not at all, I'm using the fuse based driver to back opennebula's system datastore
[16:07] <joelio> not had an issue with it - had a few with kernel based but they went away with fuse driver
[16:07] <joelio> performance is not the top of my prority for that though
[16:07] <joelio> it's not something I've heavily benched
[16:08] <nhm> joelio: ok, good to know. I've seen some random slow-downs with the kernel driver.
[16:08] <nhm> joelio: over all performance is quite good, but occasionally it stalls out for a bit.
[16:08] <nhm> glad to hear the fuse driver is working well!
[16:09] <joelio> yea, it's something I would have run a mile from in the past (fuse that is) but it seems to be quite mature and has survived to my completely non-scientific load tests
[16:09] <yanzheng> nhm,
[16:10] <yanzheng> data io stall or metadata op stall?
[16:11] <nhm> yanzheng: data IO stall. metadata possibly too. After several iterations I saw high min/max variance.
[16:11] <nhm> For numerious op types
[16:12] <yanzheng> you men
[16:12] <yanzheng> you saw "slow op" in ceph console?
[16:12] * tziOm (~bjornar@194.19.106.242) Quit (Remote host closed the connection)
[16:13] <nhm> no, sorry. I mean with something like mdtest, running through different operations like dir create, dir stat, dir delete, file create, file stat, file delete, etc. I saw high variance between runs.
[16:13] <nhm> I also saw very slow performance at times with sequential writes and very little metadata activity.
[16:13] <nhm> Though sequential write performance overall was very good.
[16:14] <yanzheng> which kernel?
[16:14] * zetheroo1 (~zeth@home.meteotest.ch) has joined #ceph
[16:14] <yanzheng> I meam which kernel version
[16:16] <nhm> yanzheng: that system was tested with 3.9. 3.5 was used previously but had major performance issues due to zone lock contention in the mm.
[16:16] * Kioob`Taff (~plug-oliv@local.plusdinfo.com) Quit (Quit: Leaving.)
[16:17] <yanzheng> 3.9 is too old ;)
[16:18] <nhm> yanzheng: When we were doing the testing it was newer. ;)
[16:19] <nhm> yanzheng: Have you got some patches into the newer kernels that help with that sort of thing?
[16:20] <yanzheng>
[16:22] * KindOne (~KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[16:22] * alphe (~alphe@0001ac6f.user.oftc.net) has joined #ceph
[16:23] <alphe> hello all :)
[16:23] <yanzheng> a few for request hangļ¼Œ none for short stall
[16:23] <alphe> I would like to try again to make s3-website share with working ssl stack
[16:24] <alphe> I noticed that it is better to use 100-continue able apache and fastcgi mod by community so I will start with that follow the documentation and I hope have it working
[16:25] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) Quit (Ping timeout: 480 seconds)
[16:25] * odyssey4me (~odyssey4m@165.233.71.2) Quit (Ping timeout: 480 seconds)
[16:25] <alphe> if I have any modifications to add to the actual documentation to whoom I can send a mail to make the changes ?
[16:26] * KindOne (~KindOne@0001a7db.user.oftc.net) has joined #ceph
[16:28] * rongze_ (~quassel@117.79.232.217) has joined #ceph
[16:30] * BillK (~BillK-OFT@58-7-52-33.dyn.iinet.net.au) Quit (Ping timeout: 480 seconds)
[16:34] * mschiff (~mschiff@p4FD7E5B8.dip0.t-ipconnect.de) has joined #ceph
[16:35] * rongze (~quassel@117.79.232.217) Quit (Ping timeout: 480 seconds)
[16:38] * mschiff (~mschiff@p4FD7E5B8.dip0.t-ipconnect.de) Quit (Remote host closed the connection)
[16:38] * alram (~alram@38.122.20.226) has joined #ceph
[16:45] * odyssey4me (~odyssey4m@41.13.216.167) has joined #ceph
[16:50] * madkiss (~madkiss@184.105.243.169) Quit (Quit: Leaving.)
[16:51] * odyssey4me2 (~odyssey4m@165.233.71.2) has joined #ceph
[16:51] * yanzheng (~zhyan@101.83.43.148) Quit (Ping timeout: 480 seconds)
[16:52] <zetheroo1> ceph or gluster? or are these two totally separate animals?
[16:53] * odyssey4me (~odyssey4m@41.13.216.167) Quit (Ping timeout: 480 seconds)
[16:54] <ircolle> zetheroo1 - you're in #ceph, what do you think people will answer? :-)
[16:54] <zetheroo1> indeed. ... I am hoping people will enlighten me to the benefits that ceph has over gluster ;)
[16:55] <ircolle> zetheroo1 - have you watched this? http://blog.techdozor.org/index.php/2013/06/11/ceph-vs-gluster-debate/
[16:55] <kraken> php is just terrible
[16:55] * xoJIog (~xoJIog@195.13.218.197) has joined #ceph
[16:56] <zetheroo1> not yet ... will do now
[16:56] <ircolle> kraken - thanks for sharing
[16:56] <joao> lol
[16:58] <xoJIog> Hi guys!
[16:58] <xoJIog> I am getting mds0 reconnect start ;mds0 reconnect success ;mds0 recovery completed several times a day
[16:58] <xoJIog> is this normal?
[16:58] * markbby (~Adium@168.94.245.4) Quit (Quit: Leaving.)
[17:01] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) has joined #ceph
[17:03] * doxavore (~doug@99-89-22-187.lightspeed.rcsntx.sbcglobal.net) has joined #ceph
[17:04] * yanzheng (~zhyan@101.83.43.148) has joined #ceph
[17:05] * markbby (~Adium@168.94.245.4) has joined #ceph
[17:05] * markbby1 (~Adium@168.94.245.4) has joined #ceph
[17:08] * gregmark (~Adium@cet-nat-254.ndceast.pa.bo.comcast.net) has joined #ceph
[17:08] * BManojlovic (~steki@91.195.39.5) Quit (Quit: Ja odoh a vi sta 'ocete...)
[17:11] * markbby (~Adium@168.94.245.4) Quit (Remote host closed the connection)
[17:14] <yo61> Am trying to use ceph-deploy
[17:14] <yo61> I get: sudo: sorry, you must have a tty to run sudo
[17:15] <alfredodeza> yo61: the user you are login in as to the remote host needs `sudo` permissions
[17:15] <yo61> It does
[17:16] <alfredodeza> yo61: can you paste the whole output from start to end so I can take a look?
[17:16] <yo61> Sure
[17:17] <yo61> https://gist.github.com/robinbowes/6282820
[17:18] <yo61> Found this: http://www.spinics.net/lists/ceph-users/msg02822.html
[17:18] <yo61> An trying commenting out requiretty in /etc/sudoers
[17:19] <yo61> OK, that fixed that
[17:19] <yo61> Now getting:
[17:19] <yo61> which: no lsb_release in (/sbin:/bin:/usr/sbin:/usr/bin)
[17:19] <yo61> /usr/lib/python2.6/site-packages/pushy/protocol/connection.py:87: DeprecationWarning: BaseException.message has been deprecated as of Python 2.6
[17:19] <yo61> ceph-deploy: The lsb_release command was not found on remote host. Please install the lsb-release package.
[17:19] <yo61> Bah
[17:19] <alfredodeza> yo61: no worries, I am working on a fix for that :)
[17:20] <alfredodeza> the next release of ceph-deploy will not bail on you like this for lsb_release :)
[17:20] <yo61> What, the lsb-release ?
[17:20] <alfredodeza> yes
[17:20] <alfredodeza> for now it is needed
[17:20] <yo61> I hate lsb with a passion
[17:21] <alfredodeza> you shouldn't care about it, right? it is just used to make sure we are dealing with the right distro/version
[17:21] <yo61> What's it needed for?
[17:21] <ircolle> yo61 - :-)
[17:21] <yo61> It pulls in a metric shit ton of un-needed packages
[17:21] <alfredodeza> yo61: when you say "install {node}" we need to make sure we are using the right packages for your OS
[17:21] * mnash (~chatzilla@vpn.expressionanalysis.com) has joined #ceph
[17:21] <alfredodeza> yum vs apt, rpm vs deb, etc...
[17:22] <alfredodeza> yo61: no, you need to install redhat-lsb-core
[17:22] <alfredodeza> we don't need everything else
[17:22] <alfredodeza> the lsb-core has a minimum set of tools that are used for ceph-deploy
[17:22] <yo61> OK
[17:23] <alfredodeza> but then again, I am working on a fix to not need this for the next release of ceph-deploy
[17:26] * ircolle (~Adium@c-67-165-237-235.hsd1.co.comcast.net) Quit (Quit: Leaving.)
[17:26] <joelio> lsb is a pretty sane way of doing it to me
[17:26] <kraken> joelio, this ain't your shell
[17:26] <yo61> redhat-lsb-core is not too bad
[17:26] * zetheroo1 (~zeth@home.meteotest.ch) has left #ceph
[17:26] <yo61> lsb used to pull in all sorts of unimaginable crap
[17:26] <kraken> yo61, this ain't your shell
[17:27] <joelio> kraken: want a botsnack?
[17:27] <alfredodeza> lol
[17:27] <yo61> mv kraken /dev/null
[17:27] <yo61> chmod 0000 kraken
[17:28] <joelio> just don't mention the .php
[17:28] <kraken> php is just terrible
[17:29] <alfredodeza> let me mute those so it is less annoying
[17:29] <yo61> gah
[17:29] <yo61> ceph-deploy want to talk to the outside
[17:30] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Remote host closed the connection)
[17:30] <yo61> I've already deployed the ceph repo and imported the key with puppet
[17:30] * xoJIog (~xoJIog@195.13.218.197) Quit (Quit: Konversation terminated!)
[17:30] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) has joined #ceph
[17:30] <alfredodeza> what do you mean by 'talk to the outsie'
[17:31] <alfredodeza> it certainly needs to install stuff for you
[17:31] <alfredodeza> get packages, keys, repos, etc...
[17:31] <alfredodeza> you can point to a proxy if you need to
[17:31] <alfredodeza> this is also documented :)
[17:31] <yo61> I have a provisioning server that contains a mirror of the ceph repo
[17:32] <yo61> The servers I'm using for ceph have no internet connectivity
[17:33] <yo61> I was hoping to build things out using puppet, but I decided I need to learn about the various moving parts of ceph first so thought I'd throw something together with ceph-deploy and work with that for a bit first
[17:33] <yo61> I've no idea why you want to rpm --import from github. anyway
[17:34] <yo61> if you're on an EL-flavour distro then there is a yum repo
[17:34] <yo61> So you can yum -y install ceph-release
[17:34] <yo61> Then yum -y install <whatever> and it will import the key
[17:34] <yo61> Oh wait, the key is not in ceph-release
[17:34] <yo61> Bah
[17:35] <yo61> What I need is a manual walkthrough
[17:35] <yo61> install this, configure this, copy this from here to there, import this, etc.
[17:35] <yo61> I've not found one yet
[17:35] <yo61> </rant>
[17:37] <alfredodeza> sorry, I am trying to process all of your input but it is a bit too much
[17:38] <alfredodeza> one thing at a time :)
[17:38] <yo61> You mean you don't auto-scale?
[17:38] <yo61> ;)
[17:39] <alfredodeza> the idea of ceph-deploy is to 1) be an easy way to install and deploy different aspects of ceph and 2) provide as much information as possible so that you can later implement the same 'actions' with chef, puppet, cfengine etc...
[17:39] <alfredodeza> if you are running ceph-deploy v 1.2 or later, you will have detailed log output for install and `mon create`
[17:39] <alfredodeza> each command is in the log output
[17:40] <yo61> I have 1.0, it seems
[17:40] <joelio> yo61: I'm inside a non-routed network too.. I just installed the ceph packages and keys by hand and ran the rest of the steps via ceph-deploy. There should be support for proxies now thouugh?
[17:41] * Cube (~Cube@et-0-30.gw-nat.bs.kae.de.oneandone.net) Quit (Quit: Leaving.)
[17:41] <alfredodeza> joelio: there is support for proxies
[17:41] <alfredodeza> not directly with ceph-deploy but with wget
[17:41] <alfredodeza> and this is also documented
[17:41] <yo61> I can do the installs manually - not a problem
[17:41] <alfredodeza> https://github.com/ceph/ceph-deploy#proxy-or-firewall-installs
[17:41] <yo61> ...if I know what to install
[17:41] <joelio> alfredodeza: as mentioned to you during the dev days, the way it was implemented diregards env vars
[17:41] <joelio> I don't know if this is now fixed or not?
[17:42] <alfredodeza> joelio: you don't need env vars
[17:42] <joelio> alfredodeza: I have env vars set for my systems automatically via puppet. There's http_proxy env vars
[17:43] <joelio> wget *should* honour those
[17:43] <joelio> as well as curl or whatever other tool
[17:43] <joelio> no need for manual .wgetrc etc.
[17:43] <alfredodeza> it should, I wasn't aware it did not
[17:43] <joelio> yep, I think because it subshells it loses the env vars
[17:43] <yo61> So, what packages do I need to install?
[17:44] <joelio> yo61: ceph ceph-common shoudl dio
[17:44] <yo61> I have no ceph-common
[17:44] * tnt_ (~tnt@212-166-48-236.win.be) Quit (Ping timeout: 480 seconds)
[17:46] <joelio> yo61: I'm on ubuntu, maybe different packages for rhel/cent - http://eu.ceph.com/docs/next/install/rpm/
[17:46] <yo61> Just ceph, I reckon
[17:46] <joelio> yep
[17:47] <yo61> Where can I get a later ceph-dpeloy tool?
[17:47] <joelio> pip
[17:47] <joelio> git
[17:47] <alfredodeza> yo61: multiple places
[17:47] <joelio> gitbuilder maybe? alfredodeza is your man
[17:47] <yo61> AM currently pulling from http://ceph.com/rpm-cuttlefish/el6/x86_64
[17:47] <alfredodeza> this is also documented (re: installation) https://github.com/ceph/ceph-deploy#installation
[17:48] <alfredodeza> yo61: have you gone through the installation section? ^ ^
[17:48] * ircolle1 (~Adium@c-67-165-237-235.hsd1.co.comcast.net) has joined #ceph
[17:49] <yo61> Not that one, no
[17:49] <yo61> Have been looking at the ceph docs
[17:50] <yo61> OK, those docs are better
[17:51] * mschiff (~mschiff@85.182.236.82) has joined #ceph
[17:54] <yo61> I installed the packages manually, using yum install ceph on all 5 nodes (well, used puppet, actually)
[17:55] <yo61> I then ran "ceph-deploy new ceph01 ceph03 ceph05"
[17:56] <yo61> Next, I ran ceph-deploy mon create ceph01 ceph03 ceph05
[17:57] * bandrus (~Adium@cpe-76-95-217-129.socal.res.rr.com) has joined #ceph
[17:57] <yo61> All good so far
[17:57] <yo61> But when I try ceph-deploy gatherkeys ... it doesn't find any keys
[17:58] <yo61> Unable to find /etc/ceph/ceph.client.admin.keyring on ['ceph01']
[17:58] <yo61> Unable to find /var/lib/ceph/bootstrap-osd/ceph.keyring on ['ceph01']
[17:58] <yo61> Unable to find /var/lib/ceph/bootstrap-mds/ceph.keyring on ['ceph01']
[17:59] * davidzlap (~Adium@ip68-5-239-214.oc.oc.cox.net) Quit (Quit: Leaving.)
[18:02] * allsystemsarego (~allsystem@5-12-241-157.residential.rdsnet.ro) Quit (Ping timeout: 480 seconds)
[18:05] * julian (~julianwa@125.69.106.188) Quit (Quit: afk)
[18:08] <yo61> Hey ho, am going to try the latest versions from ceph-testing repo
[18:08] * davidzlap1 (~Adium@ip68-5-239-214.oc.oc.cox.net) has joined #ceph
[18:10] * grepory (~Adium@50-115-70-146.static-ip.telepacific.net) has joined #ceph
[18:10] * davidzlap1 (~Adium@ip68-5-239-214.oc.oc.cox.net) Quit ()
[18:11] * madkiss (~madkiss@64.125.181.92) has joined #ceph
[18:11] <foosinn> yo61, have a serch on the web for it, there an open bug report.
[18:11] * foosinn (~stefan@office.unitedcolo.de) Quit (Quit: Leaving)
[18:12] * odyssey4me2 (~odyssey4m@165.233.71.2) Quit (Ping timeout: 480 seconds)
[18:13] * terje-_ is now known as terje-
[18:16] * tnt (~tnt@109.130.104.131) has joined #ceph
[18:18] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Ping timeout: 480 seconds)
[18:21] <alphe> ceph-deploy is a great tool
[18:21] <sage> \o/
[18:22] <alphe> yo61 you have to create one monitor first
[18:22] <alphe> then wait a few seconds
[18:22] <alphe> then create the extra ones
[18:22] <alphe> the documentation isn t clear on that topic ... or not enough precise
[18:22] <alphe> I had the same problem ...
[18:23] <alphe> you need to have first to create a single monitor that will asume the master role
[18:23] <alphe> then a few seconds later (lets make it a minute) you create the slaves ...
[18:24] <alphe> yo61 from your actual position what happends is all of them claims to be master so it sucks right ?
[18:24] <alphe> the keyring are not created correctly since everyone in the team claim to be the boss
[18:25] <alphe> yo61 to solve that issue you have to reinstall ...
[18:25] * loicd seing "test_rados.test_rados_parse_conf ... ERROR" with AttributeError: 'Rados' object has no attribute 'parse_env' . Digging.
[18:26] <loicd> when running AttributeError: 'Rados' object has no attribute 'parse_env'
[18:26] <loicd> when running qa/workunits/rados/test_python.sh
[18:26] <alphe> the best is to completly purge and then install again not forgeting to wait betwin your first mon and the next ones
[18:26] <alphe> yo61 ?
[18:26] <alphe> loicd hello sir :)
[18:27] <loicd> \o alphe
[18:27] * devoid (~devoid@130.202.135.225) has joined #ceph
[18:27] <alfredodeza> alphe: thanks for the kind words towards ceph-deploy, we are definitely trying to make it better :)
[18:27] <alphe> loicd I want to try to get right the s3 -amazon share based on the documentation with ssl working
[18:27] <alfredodeza> docs?
[18:27] <kraken> http://ceph.com/docs/master/
[18:27] <alfredodeza> yessss
[18:28] <alphe> alfredodeza it is a great tool really young but right
[18:28] <alphe> and it is a scripted tool which make possible fast code fix
[18:29] <alphe> alfredodeza yes the ceph object gateway doc
[18:29] <alphe> http://ceph.com/docs/master/radosgw/ from there down bellow
[18:30] <alphe> I already followed it (a part the custom apache and fastmod form ceph-community ) and I ended with a s3-amazon share working properly but without the SSL layer ...
[18:31] <alphe> most of s3-amazon clients don t allow the use of amazon if it isn t with a SSL layer ...
[18:32] <alphe> so I wish to tray again from scratch using the custom apache and fastcgi-mod and see if I get the whole thing workiong this time
[18:32] <alphe> Though in the /etc/apache/available-site/rgw.conf file how to there is no references of 443 port or SSL
[18:33] <alphe> so in one hand you create a ssl pipe but then your site doesn t use it ... that part of the documentation is fuzzy
[18:33] * torment1 (~torment@pool-96-228-147-37.tampfl.fios.verizon.net) Quit (Read error: Operation timed out)
[18:35] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[18:35] * ChanServ sets mode +v andreask
[18:43] * zhyan_ (~zhyan@101.83.54.175) has joined #ceph
[18:44] <alphe> alfredodeza it is a great tool really! Young but do the right way !
[18:45] <yo61> Am intermittent at the moment - about to geto n train...
[18:45] <alphe> yo61 read up the chat there is a comment of your problem
[18:46] <alphe> yo61 it is not a problem of version of ceph or ceph-deploy it is a problem of your understanding of the right way to create monitors :)
[18:46] <alphe> you need to create only one monitor for a minute
[18:47] <yo61> I am following the docs
[18:47] <alphe> then once that minute is passed you create your slave monitors (backups monitor)
[18:47] <yo61> I agree I don't understand what I'm doing
[18:47] <joelio> yea, there's key generation that needs entropy
[18:47] <joelio> docs are missing here I think
[18:47] <alphe> yo61 I know I did the same 3 weeks ago got the same problem with the keyrings and loicd guided me throught the solution
[18:48] <alphe> and the right solution is to erase the install and redo it (ceph-deploy purge etc...)
[18:49] * buck (~buck@bender.soe.ucsc.edu) has joined #ceph
[18:49] <alphe> then you reinstall and you make a pause in the install betwin the first monitor created and the other and the 1 minute time is right that way you are sure what ever hardware you use to be ok
[18:49] * torment1 (~torment@pool-71-251-73-5.tampfl.fios.verizon.net) has joined #ceph
[18:49] * tobru (~quassel@217-162-50-53.dynamic.hispeed.ch) has joined #ceph
[18:50] * yanzheng (~zhyan@101.83.43.148) Quit (Ping timeout: 480 seconds)
[18:50] <alphe> joelio +10 but it is not so much of a big deal just saying ceph-deploy create mon ceph01 then pause 1 minute then ceph-deploy create mon ceph{02-03} would solve the problem
[18:50] * roald (~oftc-webi@87.209.150.214) has joined #ceph
[18:51] <joelio> alphe: agreed, simple doc tweak.. in our internal docs I added a big 'WAIT FOR A MINUTE - KEYS ARE BEING CREATED'
[18:51] <joelio> as it cauight me out a few times
[18:51] * markbby1 (~Adium@168.94.245.4) Quit (Quit: Leaving.)
[18:51] <alphe> yo61 to resume you create 3 monitors that all three will claim to be the master monitor so they will claim right to create and manage the keyrings ...
[18:52] <alphe> so you will get a keyring issue when you try to gather the keyrings since everyone is he boss and give a diffrent key
[18:52] <yo61> Sure would be good to have high-level docs describing the steps needed to create a new cluster from scratch
[18:53] <alphe> you need one boss and then backup boss that will ask to the boss the only unique keyring
[18:54] <alphe> so when you gateher the keyrings everyone is on the same page and return a single keyring that everyone shares
[18:54] <alphe> in the ceph cluster
[18:54] <yo61> Yeah, I get that now :)
[18:55] <alphe> so from now on you need to clean you ceph cluster uninstall all purge all and then restart ...
[18:56] <alphe> to clean all the right way is :
[18:56] <alphe> ceph-deploy -v purge osd01 &
[18:56] * ishkabob (~c7a82cc0@webuser.thegrebs.com) has joined #ceph
[18:56] <ishkabob> hey guys, can i add placement groups to a pool that is currently being written to?
[18:56] * markbby (~Adium@168.94.245.4) has joined #ceph
[18:56] <alphe> ceph-deploy -v purge osd01 & and you repleat that command on each of your node
[18:57] <alphe> yo61 in your case it is ceph-deploy -v purge ceph{01-03} &
[18:57] <alphe> then you purge the data
[18:57] <alphe> ceph-deploy -v purge-data ceph{01-03}
[18:58] <alphe> you can complete with an
[18:58] <alphe> extra step ceph-deploy uninstall ceph{01-03}
[18:58] <alphe> so that cleansen all your nodes ...
[18:59] <yo61> k
[18:59] <alphe> then you start from begining you install ceph on you admin node
[18:59] <yo61> What iptables ports are required?
[18:59] <alphe> yo61 the same as you allready have
[19:00] <yo61> ?
[19:00] <alphe> since ceph-deploy was working before it should use the same communication pipes to imput commands
[19:00] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Ping timeout: 480 seconds)
[19:01] <yo61> I mean for ceph itself - the mon communication, OSDs, etc.
[19:01] <alphe> you could use ceph-deploy previously to install I guess there is no other problems ... not sure since I kill iptable and selinux at os install stage to not have them bothering me...
[19:02] <alphe> jeolio wait for a minute without the meaning is hard to get ... Wait for a minute a master have to be created or you will have concurent claims to be master and that will bring keyring problems at gathering stage ...
[19:03] <alphe> yo61 this doc should answer your ask :
[19:03] <alphe> http://eu.ceph.com/docs/wip-msgauth/config-cluster/ceph-conf/
[19:03] <alphe> iptables -A INPUT -m multiport -p tcp -s 192.168.1.0/24 --dports 6789,6800:6803 -j ACCEPT
[19:04] <alphe> 6789 is the monitor port that much I know since it is the port you use to contact the monitor to mount with ceph-fuse
[19:05] <alphe> i imagine 6800 and 6803 are for osd(s) and md(s)
[19:05] <yo61> That's a range of ports
[19:05] <alphe> Monitors listen on port 6789, while metadata servers and OSDs listen on the first available port beginning at 6800.
[19:05] <yo61> So, 6800,6801,6802,6803
[19:05] <alphe> Monitors listen on port 6789, while metadata servers and OSDs listen on the first available port beginning at 6800.
[19:05] <alphe> that is what the doc says I should read more it is all said :)
[19:14] * mozg (~andrei@host217-46-236-49.in-addr.btopenworld.com) Quit (Ping timeout: 480 seconds)
[19:20] <yo61> OK, doing a new install
[19:20] <yo61> Or rather, new deploy
[19:20] <yo61> ceph-deploy --overwrite-conf --cluster cloud new ceph01 ceph03 ceph05
[19:21] <yo61> ceph-deploy --overwrite-conf --cluster cloud mon create ceph01
[19:21] <yo61> sleep 60
[19:21] * lightspeed (~lightspee@81.187.0.153) has joined #ceph
[19:21] <yo61> How can I tell if the mon is ready ?
[19:22] * LPG (~LPG@c-76-104-197-224.hsd1.wa.comcast.net) has joined #ceph
[19:23] * devoid (~devoid@130.202.135.225) Quit (Quit: Leaving.)
[19:23] <dmick> yo61: if it answers ceph -s, it's ready
[19:23] <dmick> if it hangs, you'll have to get more creative
[19:24] <yo61> Hmm
[19:24] <dmick> er....hang on, it won't answer if you only started one
[19:24] <yo61> /etc/init.d/ceph: ceph conf /etc/ceph/ceph.conf not found; system is not configured.
[19:24] <dmick> yeah, you'll have to use --cluster cloud on everything
[19:24] <dmick> but start the other two before you try
[19:24] <yo61> I did
[19:24] <yo61> Oh
[19:25] <yo61> Looks like it's not writing to the right place
[19:25] <dmick> wait...who's running /etc/init.d/ceph?
[19:25] <yo61> write cluster configuration to /etc/ceph/{cluster}.conf
[19:25] <yo61> This is ceph-deploy
[19:25] <dmick> the error message was from a ceph-deploy command?
[19:25] <yo61> No, I ran he same commands as ceph-deploy was running
[19:25] <yo61> To see why the mon had not started
[19:26] <dmick> ??!
[19:26] <yo61> I ran: ceph-deploy --overwrite-conf --cluster cloud mon create ceph01
[19:26] <yo61> And looked on ceph01 to see if it had started
[19:26] <yo61> It hadn't
[19:26] <dmick> "looked" how?
[19:26] <yo61> ps -ef | grep
[19:26] <yo61> on ceph01
[19:27] <dmick> ok
[19:27] <yo61> So I then ran the same command that ceph-deploy used
[19:27] <yo61> And got that error
[19:27] <yo61> Looks like I need to tell it use cloud.conf not ceph.conf
[19:27] <dmick> it's very difficult to do the same thing that ceph-deploy does
[19:28] <dmick> I suspect some small detail was wrong in the recreation, but, we should figure out what the state of the mon is there
[19:28] <yo61> Er, well it said: Running command: /sbin/service ceph start mon.ceph01
[19:28] <dmick> I have to run to a meeting, but I'd start by looking at the mon logs
[19:28] <yo61> So I ran the same command
[19:28] <alphe> ceph-deploy write in the local dir you run the command from
[19:29] <alphe> you have a ceph.log in that local dir that gives you hint of what is happening
[19:29] <yo61> Yes
[19:29] <yo61> Same thing
[19:29] <alphe> you can read /var/log/ceph.log too or something like that
[19:29] <alphe> you can use ceph-deploy -v for verbose too
[19:30] * doxavore (~doug@99-89-22-187.lightspeed.rcsntx.sbcglobal.net) Quit (Ping timeout: 480 seconds)
[19:30] <alphe> which is pretty usefull when you cascade commandes to all your nodes
[19:30] <yo61> It looks like the init script is trying to to read ceph.conf, not cloud.conf
[19:31] <alphe> you allways have to be in the local dir that has your ceph.log ceph.conf etc files ...
[19:31] <alphe> to transfer those files to you /etc/ceph you should use from your local dir ceph-deploy create admin once all other tasks are done
[19:32] <alphe> since you will create alots of new entries to ceph-deploy you don t need to populate it to your admin server /etc/ceph every time ...
[19:33] <alphe> so once you have the monitors the disks the osds the mds running you can use ceph-deploy admin to write the files from local dir to /etc/ceph ...
[19:33] <alphe> though it is not mandatory ...
[19:33] <alphe> you can cp them etc...
[19:34] <yo61> Gott a jump off the train now
[19:34] <alphe> more details on ceph-deploy admin and config file distrubution (pull push )
[19:34] <alphe> http://ceph.com/docs/master/rados/deployment/ceph-deploy-admin/
[19:34] <alphe> ok :)
[19:35] * portante|afk is now known as portante
[19:35] <alphe> 1 portante :)
[19:36] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) Quit (Remote host closed the connection)
[19:36] * mschiff (~mschiff@85.182.236.82) Quit (Ping timeout: 480 seconds)
[19:39] * portante wonders when the Count's laugh, lightning and thunder will follow ... ah ah ah!
[19:39] <portante> alphe: ;)
[19:39] * xmltok_ (~xmltok@pool101.bizrate.com) Quit (Remote host closed the connection)
[19:39] * xmltok (~xmltok@relay.els4.ticketmaster.com) has joined #ceph
[19:40] * xmltok_ (~xmltok@pool101.bizrate.com) has joined #ceph
[19:40] * xmltok_ (~xmltok@pool101.bizrate.com) Quit (Remote host closed the connection)
[19:40] * xmltok (~xmltok@relay.els4.ticketmaster.com) Quit (Read error: Connection reset by peer)
[19:40] <joelio> I should be able to check if rbd_cache is enabled via an asok right?
[19:40] * xmltok (~xmltok@relay.els4.ticketmaster.com) has joined #ceph
[19:41] * xmltok_ (~xmltok@pool101.bizrate.com) has joined #ceph
[19:41] * xmltok_ (~xmltok@pool101.bizrate.com) Quit (Remote host closed the connection)
[19:41] * xmltok_ (~xmltok@pool101.bizrate.com) has joined #ceph
[19:43] * doxavore (~doug@99-89-22-187.lightspeed.rcsntx.sbcglobal.net) has joined #ceph
[19:44] * mschiff (~mschiff@85.182.236.82) has joined #ceph
[19:45] <gregaf> sagewk: did you want to talk about caching pools and pg versions now?
[19:45] <gregaf> or sage?
[19:45] <sage> let's do it when there is a whiteboard.. tomorrow?
[19:46] <gregaf> well it's the task I have to do right now :p
[19:46] <alphe> damn internet have no whiteboard on it why is that ?!
[19:46] <alphe> hehehe ...
[19:46] <gregaf> so I can make up 8 hours worth of work and have it be useful, but I'd rather not
[19:48] <mikedawson> joelio: ceph --admin-daemon /var/run/ceph/ceph-osd.0.asok config show | grep rbd_cache
[19:48] * xmltok (~xmltok@relay.els4.ticketmaster.com) Quit (Ping timeout: 480 seconds)
[19:49] <gregaf> all right, I'm just going to put the intelligence in the objecter, it'll be a little more branchy than I like but not complicated
[19:52] * tnt_ (~tnt@109.130.102.13) has joined #ceph
[19:52] <alphe> gregaf hum if the wait means you will have something that will not implies you to loose more and more and more time next it is probably worth the way
[19:53] <alphe> quick and dirty is not meant to stay long so soon or later it has to be long and cleany ...
[19:54] <Kioob> Is there sort of "memory storage" in Ceph ?
[19:54] * tnt (~tnt@109.130.104.131) Quit (Ping timeout: 480 seconds)
[19:54] <gregaf> nah, I'm pretty sure it has to be this way I just wanted to draw it with somebody and make sure, and sjust agrees with me that we can't do better ;)
[19:54] <gregaf> Kioob: not unless you're using ramdisks :p
[19:54] <alphe> kioob what do you mean sharing RAM betwin nodes in ceph cluster ?
[19:54] <Kioob> gregaf: good idea ! :D
[19:55] <alphe> ceph isn t beowolf os :)
[19:55] <Kioob> I like the way Ceph manage consistency between multiple hosts, and I'm searching to use similar things to store volatile data :D
[19:57] <alphe> kioob hum ramdisk :) see that ceph cluster is it a whole amazing 1TB of all the ram available everywhere here totally useless but so great !
[19:57] <Kioob> :)
[19:58] <alphe> kioob too many unused, bored and lonely GB or RAM now by days :)
[19:59] * devoid (~devoid@130.202.135.225) has joined #ceph
[19:59] <alphe> kioob beowulf would gather the ram of all your nodes in the beowulf cluster but not make them available for data storage as filesystem ...
[20:00] <Kioob> Thought TMEM you can aggregate RAM via network :D
[20:01] <alphe> and bring your ram to the world of the amazing Gigabit ethernet speed ?
[20:01] <alphe> hehehe ...
[20:01] <Kioob> yep !
[20:07] <Kioob> In fact my "problem" is storing the VM / RBD configs in a DB, which will probably be less safe than Ceph. So I'm thinking to a way to use that, but I should probably use a safe DB setup, instead of trying to put that in Ceph
[20:08] <acaos_> if I have ceph journals on an SSD, is there a benefit to storing them on partitions/block devices rather than files on a filesystem?
[20:08] * acaos_ is now known as acaos
[20:09] <nhm> acaos_: one less thing to get in the way basically
[20:09] <acaos> yeah, I know that, I was more wondering if there were significant performance implications
[20:09] <nhm> acaos: also you can just partition what you need and leaving some free unpartitioned space may help with wear-levelling depending on the drive.
[20:09] <acaos> I don't -think- so because 'journal dio' should bypass that
[20:10] * jbd_ (~jbd_@2001:41d0:52:a00::77) has left #ceph
[20:10] <nhm> acaos: I suspect you probably won't see that much difference to be honest.
[20:11] <acaos> ok, thank you, I didn't think I would either
[20:11] * Cube (~Cube@88.128.80.12) has joined #ceph
[20:14] <acaos> on a separate note, is there any hope of breaking out the librados includes/etc from ceph-devel into a librados2-devel package (which ceph-devel could then depend on)? That way it wouldn't be necessary to install all of ceph to do librados development
[20:15] <nhm> acaos: interesting, you might want to raise that on the mailing list.
[20:15] <acaos> ok, I will
[20:16] <acaos> (I just recently had the fun experience of having a build server yank in the entire ceph infrastructure to build a librados program)
[20:16] <nhm> ugh
[20:19] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[20:19] * ChanServ sets mode +v andreask
[20:20] * markbby (~Adium@168.94.245.4) Quit (Remote host closed the connection)
[20:21] * markbby (~Adium@168.94.245.4) has joined #ceph
[20:24] * dpippenger (~riven@tenant.pas.idealab.com) has joined #ceph
[20:34] * carif (~mcarifio@pool-96-233-32-122.bstnma.fios.verizon.net) has joined #ceph
[20:37] * dpippenger1 (~riven@tenant.pas.idealab.com) has joined #ceph
[20:37] * dpippenger (~riven@tenant.pas.idealab.com) Quit (Read error: Connection reset by peer)
[20:38] * carif (~mcarifio@pool-96-233-32-122.bstnma.fios.verizon.net) Quit ()
[21:05] <sjust> k
[21:09] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) Quit (Quit: Leaving.)
[21:10] * sstan (~chatzilla@dmzgw2.cbnco.com) has joined #ceph
[21:15] * devoid (~devoid@130.202.135.225) Quit (Quit: Leaving.)
[21:16] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) Quit (Ping timeout: 480 seconds)
[21:16] * devoid (~devoid@130.202.135.225) has joined #ceph
[21:20] * wschulze (~wschulze@cpe-69-203-80-81.nyc.res.rr.com) has joined #ceph
[21:23] * sstan (~chatzilla@dmzgw2.cbnco.com) Quit (Quit: ChatZilla 0.9.90.1 [Firefox 19.0/2013021500])
[21:24] * rongze (~quassel@117.79.232.249) has joined #ceph
[21:30] * rongze_ (~quassel@117.79.232.217) Quit (Ping timeout: 480 seconds)
[21:42] <alphe> kioob the problem I see with a ramdisk volatile ceph storage is that you will have a MDS that will record data distribution that doesn t exists on next reboot ...
[21:43] <alphe> you have to start the volatile cluster all together and stop it at once
[21:43] * zhyan__ (~zhyan@101.83.119.69) has joined #ceph
[21:48] * mozg (~andrei@host109-151-35-94.range109-151.btcentralplus.com) has joined #ceph
[21:50] * zhyan_ (~zhyan@101.83.54.175) Quit (Ping timeout: 480 seconds)
[21:56] * alram (~alram@38.122.20.226) Quit (Ping timeout: 480 seconds)
[22:02] * ishkabob (~c7a82cc0@webuser.thegrebs.com) Quit (Quit: TheGrebs.com CGI:IRC (Ping timeout))
[22:11] <sage> sjust: another unfound objects hang... nuke it or do you want to look?
[22:12] <sjust> i want to take a look
[22:12] * jeffhung (~jeffhung@60-250-103-120.HINET-IP.hinet.net) Quit (Ping timeout: 480 seconds)
[22:12] <sage> ubuntu@teuthology:/a/teuthology-2013-08-20_01:00:13-rados-next-testing-basic-plana/2645
[22:21] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) has joined #ceph
[22:21] <nhm> sage: https://docs.google.com/a/inktank.com/document/d/11yj_JtEW5E4nva6a1XxhVctmK2ozYhYbJ77D7il5d74/edit
[22:23] * alram (~alram@cpe-76-167-50-51.socal.res.rr.com) has joined #ceph
[22:23] <sage> zackc: Tamil: alfredodeza: nhm: can we standardize on either - or _ in teuthology yaml? i fear the mix of both is going to end in pain
[22:23] * sage likes -, but honestly who cares
[22:24] <alfredodeza> hear hear for normalization
[22:24] <Tamil> sage: - makes sense
[22:25] <nhm> sage: doesn't matter to me. Teuthology is way too random already imho.
[22:25] <sage> nhm: one ocmment int eh doc
[22:25] <zackc> sage: sorry, context?
[22:26] <sage> os_type: foo vs task ceph-fuse, interactive-on-error: true, etc.
[22:26] <zackc> ohhh right.
[22:26] <joelio> mikedawson: great, thanks!
[22:27] <zackc> sage: well, "foo_bar" is actually valid in python whereas "foo-bar" is not
[22:27] * zackc votes for _
[22:27] <sage> works for me!
[22:31] * nhorman (~nhorman@hmsreliant.think-freely.org) Quit (Quit: Leaving)
[22:32] <alphe> no - and no _
[22:32] <alphe> :)
[22:32] <alfredodeza> nooooooooooooooo
[22:32] <alphe> oneboringsinglelongphraseismyfunctionname :P
[22:32] <alfredodeza> kraken: loljava
[22:32] <kraken> AsynchronousJDBCConfigurationExtractionFactory
[22:33] <Tamil> sage: oh i thought we are talking about the values part, zack is right. we could stick with '_' as '-' is not accepted in python
[22:33] <gregaf> oh no, kraken: loljava
[22:33] <kraken> RecursiveCategoryMutationApplet
[22:33] <nhm> none of this affects performance testing, I care not.
[22:33] <sjust> kraken: lolc++
[22:33] <alfredodeza> ha!
[22:33] <alfredodeza> good try
[22:33] <sjust> should just segfault
[22:33] <alfredodeza> LOL
[22:34] <sage> lol
[22:34] <alphe> alfkrakenisabotanddontevenknowsitsopleaseignoreit !
[22:35] <alphe> kraken: lolrebol
[22:35] <alphe> kraken is right there is no lol about rebol :)
[22:35] * alphe sights
[22:35] * dmick has no idea what rebol is
[22:36] * tobru (~quassel@217-162-50-53.dynamic.hispeed.ch) Quit (Ping timeout: 480 seconds)
[22:37] <alphe> rebol is a minimalistic syntax scripting language created by Carl Sassenrath in 1998 (author of amiga OS)
[22:37] <nhm> fun
[22:38] <alphe> to show a window with a button in rebol you do rebol [] view layout [ btn green "a BUTTON" [ quit] ]
[22:39] <alphe> and so that displays a window in your os with a green button in it with a BUTTON written in it and the button action or callback is the quit function which quits
[22:39] <alphe> you are a pro to rebol now a thousand bucks please :)
[22:39] <alphe> hehehehe :)
[22:40] <alphe> more seriously rebol is opensource now under bsd licence and there is tons of things to do with it for it
[22:40] <alphe> rebol.com for a starter
[22:40] <nhm> alphe: I'll pay you in "ET Lizard Creature and Me" 2600 cartridges.
[22:41] <alphe> then stackoverflow rebol then you ccan get cheyenne! which is a web server written fully in rebol using async technologies
[22:41] <alphe> oh and rebol is only 700 kb sized
[22:42] <dmick> sounds like a rebol without a cause to me :)
[22:42] <janos> haha
[22:42] * rongze (~quassel@117.79.232.249) Quit (Read error: Connection reset by peer)
[22:42] * rongze (~quassel@106.120.176.78) has joined #ceph
[22:42] <nhm> point dmick
[22:42] <alphe> In my wildest dream I will do a cheyenne! based web_UI in rebol to manage and share ceph cluster through s3 amazon
[22:43] <alphe> or something like that :)
[22:43] <dmick> dmick: no pun too stupid
[22:43] <janos> amen
[22:47] <alphe> it was at first a scripting language for IOS (INTERNET OPERATING SYSTEM) but then it was extended and have a life on its own now
[22:48] * jeffhung (~jeffhung@60-250-103-120.HINET-IP.hinet.net) has joined #ceph
[22:51] * jeffhung (~jeffhung@60-250-103-120.HINET-IP.hinet.net) Quit ()
[22:51] * jeffhung (~jeffhung@60-250-103-120.HINET-IP.hinet.net) has joined #ceph
[22:53] <loicd> http://pastebin.com/3btJLcmw teuthology workunit rados/test_python.sh fails with
[22:53] <loicd> AttributeError: 'Rados' object has no attribute 'parse_env'
[22:53] <loicd> when I run nosetest manually from source it works
[22:54] <dmick> loicd: python-ceph is out of date
[22:54] <loicd> I'm still new to teuthology and unsure if I'm doing something wrong
[22:54] <loicd> dmick: how do you mean ?
[22:58] <dmick> there is a package named python-ceph that does not contain the latest software? :)
[22:58] <dmick> (parse_env was added in the run up to dumpling)
[22:58] <dmick> (by me)
[22:59] <dmick> this is the version on your machine-under-test, of course
[22:59] * loicd digging
[23:03] <loicd> dmick: I must be missing something. python-ceph is created by debian/control and teuthology should pick it from the build http://gitbuilder.sepia.ceph.com/gitbuilder-ceph-deb-precise-amd64-basic/log.cgi?log=9cea33ab39f5f6bfa20d99f10c94d38f9948d60d . And since
[23:03] <loicd> def test_rados_parse_conf():
[23:03] <loicd> with Rados(conffile='', rados_id='admin') as rados:
[23:03] <loicd> rados.parse_env()
[23:03] <loicd> rados.parse_env('FOO_DOES_NOT_EXIST_BLAHBLAH')
[23:03] <loicd> pass
[23:03] <loicd> works from soruce, I assume it means the python module knows parse_env
[23:03] <dmick> sigh, sorry
[23:03] <loicd> dmick: what am I missing ?
[23:03] <dmick> yes, it's not python-ceph, it's (apparently) librados2
[23:04] <dmick> should have read that more carefully, apologies
[23:04] <loicd> dmick: no worries, I very much appreciate you being there while I'm totaly confused :-)
[23:04] <loicd> s/there/here/ ( typical french mistake )
[23:05] <dmick> well, actually, LA is "there" from where you are, so it's perfectly accurate
[23:06] <dmick> IRC is a virtual "here"
[23:06] <loicd> and LA means there in french ;-)
[23:06] <dmick> heh
[23:07] <alphe> ET lizard creatures and me :)
[23:07] <alphe> hehehe ...
[23:08] * roald (~oftc-webi@87.209.150.214) Quit (Quit: Page closed)
[23:10] <loicd> http://gitbuilder.ceph.com/ceph-deb-precise-x86_64-basic/ref/master/pool/main/c/ceph/librados2_0.67-254-g9cea33a-1precise_amd64.deb
[23:10] * DarkAceZ (~BillyMays@50.107.55.36) Quit (Ping timeout: 480 seconds)
[23:10] <loicd> is dated 20-Aug-2013 10:58
[23:11] <loicd> and that's what I have :
[23:11] <loicd> ii librados2 0.67-254-g9cea33a-1precise
[23:11] <dmick> got me then. if the test is actually executing with that python and that librados, it should not be getting that error
[23:11] * DarkAce-Z (~BillyMays@50.107.55.36) has joined #ceph
[23:12] <dmick> any stray LD_LIBRARY_PATH settings?...
[23:12] <dmick> sigh.
[23:12] <dmick> once more I have to apologize.
[23:12] * sage hates autotools
[23:12] <dmick> parse_env was *just* added by sage
[23:13] <dmick> yesterday
[23:13] <dmick> so it really is python-ceph, and it needs to be as new as 24 hours old
[23:14] <loicd> dmick: indeed. But all seems in order, it's a mystery :-)
[23:16] * BillK (~BillK-OFT@58-7-52-33.dyn.iinet.net.au) has joined #ceph
[23:18] <joao> <loicd> and LA means there in french ;-) <- funny enough, in Portuguese as well
[23:19] <loicd> joao: :-)
[23:20] * carif (~mcarifio@pool-96-233-32-122.bstnma.fios.verizon.net) has joined #ceph
[23:24] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) Quit (Ping timeout: 480 seconds)
[23:25] * alfredodeza (~alfredode@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Remote host closed the connection)
[23:29] <sage> yehudasa_: can you help me with a linker problem?
[23:31] <alphe> sage don t be so hatefull autotools are great think of it ... how would be the world if you had to rely on vi to manually create your makefiles each time you want to compile ...
[23:31] <yehudasa_> sage: sure
[23:31] <alphe> the only problem is that autotools are stuck in 1997 :)
[23:31] <janos> think how intimately awesome you'd be at it if you did have to make them in vi each time! ;)
[23:32] <ron-slc> Hello all, general question. When a deep-scrub is being performed, is it possible to use DIRECT_IO, to avoid pushing more useful data out of cache?
[23:32] * kraken (~kraken@c-24-131-46-23.hsd1.ga.comcast.net) Quit (Ping timeout: 480 seconds)
[23:32] <alphe> "dang how I write text in the file" and vi ./myfiles is the most easy part of it
[23:32] <sage> yehudasa_: https://github.com/liewegas/ceph/commits/wip-crc
[23:32] <sage> and make, say, ceph-syn
[23:33] <sage> i can't make it link to the function defined in the common/crc32c_intel_fast_asm.S file :/
[23:33] * aliguori (~anthony@32.97.110.51) Quit (Remote host closed the connection)
[23:33] <alphe> that reminds me I compiled apache2 cephtuned without reading the README file ...
[23:33] <alphe> what a bad me ...
[23:33] <sage> nm on the assembled object shows
[23:33] <sage> 0000000000000000 T crc32_iscsi_00
[23:33] <sage> but nm on the libcommon.a shows:
[23:33] <sage> U crc32_iscsi_00
[23:33] <alphe> readme seems to be the default one ...
[23:34] <sage> (apt-get install yasm, btw)
[23:35] * kyann (~kyann@did75-15-88-160-187-237.fbx.proxad.net) has joined #ceph
[23:35] <yehudasa_> hmm
[23:35] <yehudasa_> .. still compiling
[23:36] <yehudasa_> ahrm .. configuring
[23:36] <yehudasa_> ./configure: line 12123: syntax error near unexpected token `fi'
[23:36] <yehudasa_> ./configure: line 12123: `fi'
[23:38] <sage> oh i forgot the m4 rule
[23:38] <sage> pull
[23:38] <dmick> yasm. O_o
[23:39] <sage> intel: "It is very portable, kept up-to-date with new instructions and allows building effortlessly on unix and windows. Is it possible that you can use yasm instead of nasm?"
[23:41] <yehudasa_> sage: is it using the flipped x86 instructions?
[23:41] <sage> flipped?
[23:41] <dmick> you mean at&t syntax vs Intel syntax?
[23:41] <yehudasa_> yep, what he said
[23:41] <sage> intel
[23:41] <sage> part of the reason for yasm
[23:42] * doxavore (~doug@99-89-22-187.lightspeed.rcsntx.sbcglobal.net) Quit (Ping timeout: 480 seconds)
[23:43] * andreask (~andreask@h081217135028.dyn.cm.kabsi.at) has joined #ceph
[23:43] * ChanServ sets mode +v andreask
[23:51] <yehudasa_> sage: need to take off for an hour or so, but we can take autotools out of the equation, can be reproduced with this: http://pastebin.com/yFjHUnrZ
[23:51] * doxavore (~doug@99-89-22-187.lightspeed.rcsntx.sbcglobal.net) has joined #ceph
[23:51] <yehudasa_> and then just g++ crc.o main.cc
[23:51] <sage> k thanks
[23:52] <alphe> To build mod_fastcgi as a project you'll need M$ VC++ 6.0 ... funny world ... I doubt there is such thing for my ubuntu
[23:52] <yehudasa_> sage: I'd try to play with all the global definitions in the .S file, maybe remove it completely
[23:52] <sage> blah, well it works with that test file :)
[23:52] <sage> so the asm .o is okay
[23:52] <yehudasa_> hmm.. didn't work for me
[23:53] <sage> gcc crc.c -o crc common/crc32c_intel_fast_asm.yasmo
[23:53] <sage> oh.. c++!
[23:53] <sage> fails with g++
[23:53] <yehudasa_> right ...
[23:53] <yehudasa_> extern "C" ...
[23:54] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) Quit (Ping timeout: 480 seconds)
[23:58] * doxavore (~doug@99-89-22-187.lightspeed.rcsntx.sbcglobal.net) Quit (Read error: Connection reset by peer)
[23:59] * torment1 (~torment@pool-71-251-73-5.tampfl.fios.verizon.net) Quit (Ping timeout: 480 seconds)

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