#ceph IRC Log


IRC Log for 2012-05-10

Timestamps are in GMT/BST.

[0:01] * s[X] (~sX]@eth589.qld.adsl.internode.on.net) has joined #ceph
[0:06] * sjustlaptop (~sam@24-205-39-1.dhcp.gldl.ca.charter.com) has joined #ceph
[0:06] * dennisj_ is now known as dennisj
[0:22] * sjustlaptop (~sam@24-205-39-1.dhcp.gldl.ca.charter.com) Quit (Quit: Leaving.)
[0:39] * aa (~aa@r200-40-114-26.ae-static.anteldata.net.uy) Quit (Ping timeout: 480 seconds)
[0:42] * LarsFronius (~LarsFroni@95-91-243-252-dynip.superkabel.de) Quit (Quit: LarsFronius)
[0:43] * S0NiC_ (~john@p54A30A2A.dip0.t-ipconnect.de) has joined #ceph
[0:43] * gregorg_taf (~Greg@ Quit (Read error: Connection reset by peer)
[0:43] * gregorg (~Greg@ has joined #ceph
[0:50] * S0NiC (~john@p54A307DE.dip0.t-ipconnect.de) Quit (Ping timeout: 480 seconds)
[0:57] * sjustlaptop (~sam@24-205-39-1.dhcp.gldl.ca.charter.com) has joined #ceph
[1:12] * jantje_ (jan@paranoid.nl) has joined #ceph
[1:13] * jantje (jan@paranoid.nl) Quit (Ping timeout: 480 seconds)
[1:29] * yehudasa_bg (~yehudasa_@mb82736d0.tmodns.net) has joined #ceph
[1:30] <yehudasa_bg> sjust: sjustlaptop: around?
[1:30] * jlogan (~chatzilla@2600:c00:3010:1:3401:eed2:6946:d8ee) Quit (Ping timeout: 480 seconds)
[1:30] <sjustlaptop> yup
[1:31] <yehudasa_bg> there are a few pgs on congress that are down+peering, though the osds that they're mapped to are up
[1:31] <sjustlaptop> what has happened recently to congress?
[1:31] <yehudasa_bg> I think there were some networking issues, they might have upgraded some switches
[1:32] <sjustlaptop> are all osds up?
[1:32] <yehudasa_bg> yes
[1:32] <yehudasa_bg> maybe related to recent stuff that sage did?
[1:32] <yehudasa_bg> the no flapping fixes
[1:32] <sage> 806 ceph health detail
[1:32] <sage> 807 ceph pg 5.f5e3 query
[1:32] <sage> 808 ceph osd dump | grep 1009
[1:33] <sage> need to bring that osd up
[1:33] <sjustlaptop> sorry, which osd?
[1:34] <sage> 1009
[1:34] <sage> just started it
[1:34] <sjustlaptop> ok, that should bring the pgs back, right?
[1:35] <yehudasa_bg> sage: any reason why osd was down?
[1:36] <sage> msg/SimpleMessenger.cc: 855: FAILED assert(existing->state == STATE_CONNECTING || existing->state == STATE_OPEN)
[1:36] <sage> blah
[1:36] <yehudasa_bg> is that a known bug?
[1:36] <sage> HEALTH_WARN 20 pgs recovering; recovery 96/1660614 degraded (0.006%)
[1:36] <sage> old bug, yeah. haven't seen it in a while tho
[1:36] <sage> one motivator for the much-talked-about messenger rewrite..
[1:37] <sage> ok now
[1:37] <sage> fwiw, the diagnosis procedure for this is documented.
[1:37] <sage> ceph health detail -- tells you which pgs are down+peering
[1:37] <sage> ceph pg <pgid> query --- tells you why
[1:38] <sage> in this case,
[1:38] <sage> "peering_blocked_by": [
[1:38] <sage> { "osd": 1009,
[1:38] <sage> "current_lost_at": 0,
[1:38] <sage> "comment": "starting or marking this osd lost may let us proceed"}]},
[1:38] <yehudasa_bg> yeah, for some reason I thought all osds were up, misread the ceph -s output.
[1:39] <sage> all the in ones were up.. it down+out is normally not problematic
[1:47] <yehudasa_bg> sage: does it make sense to extend the output there, e.g., in ceph -s note when there are pgs mapped to down osds
[1:48] <yehudasa_bg> or, specify the specific problematic osds
[1:49] <sjustlaptop> yehudasa_bg: it's not that the pgs are mapped to down osds, it's that the down osd is blocking peering
[1:50] <sage> which is what 'down' means (peering can't complete because of down osds)
[1:52] <sjustlaptop> sage: on a related note, how did one down osd block peering on a pg? There was an interval where the pg might have gone active with only 1009 in the acting set?
[1:53] <sage> apparently.. but the past intervals aren't included in the pg query results, so not sure
[1:53] <yehudasa_bg> sage: sjustlaptop: on congress there are more than 200 down osds, most of are (for now) permanently down. They don't pose any issue to the system's health
[1:53] <sage> carl was going to wipe those out (ceph osd rm N) at one point but didn't get around to it. doing that will make it easier to see things that are down but shouldn't be.
[1:54] <yehudasa_bg> well.. we shouldn'
[1:54] <yehudasa_bg> blah
[1:54] <sage> currently a failed disk means a 5 minute interval of down+in, and then down+out and some recovery time, and then everything is "ok" again. easy to not notice, depending on the monitoring..
[1:54] <yehudasa_bg> we shouldn't assume this is the exception
[1:55] <sjustlaptop> I mean that we might want to add logic to specify a minimum acting set size for going active in order to decrease the odds of a pg stuck down due to one osd
[1:56] <sage> i think that would be worse for overall availability.. you can't use your data now because if you do there's a chance that later you won't be able to
[1:56] <sjustlaptop> it would definitely be worse for availability, but one probably wouldn't expect that loosing one osd could render a 3-replica pg unusable
[1:57] <sage> yeah
[1:57] <sjustlaptop> meh, something to think about later anyway
[2:02] <yehudasa_bg> sage: sjust: in any case I think we do need to have some extra tagging to osds that are down and it affects system's availability.
[2:03] <yehudasa_bg> we shouldn't count on users to garbage collect old none-relevant osds
[2:03] * CristianDM (~CristianD@host217.190-230-240.telecom.net.ar) has joined #ceph
[2:09] * joao (~JL@aon.hq.newdream.net) Quit (Quit: Leaving)
[2:09] * yehudasa_bg (~yehudasa_@mb82736d0.tmodns.net) Quit (Read error: Connection reset by peer)
[2:12] * yehudasa_bg (yehudasa_b@static-66-14-234-139.bdsl.verizon.net) has joined #ceph
[2:13] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) Quit (Ping timeout: 480 seconds)
[2:42] * bchrisman (~Adium@ Quit (Quit: Leaving.)
[2:42] * yoshi (~yoshi@p3167-ipngn3601marunouchi.tokyo.ocn.ne.jp) has joined #ceph
[3:44] * yehudasa_bg (yehudasa_b@static-66-14-234-139.bdsl.verizon.net) Quit (Ping timeout: 480 seconds)
[3:44] * yehudasa_bg (yehudasa_b@static-66-14-234-139.bdsl.verizon.net) has joined #ceph
[3:53] * chutzpah (~chutz@ Quit (Quit: Leaving)
[4:00] * yehudasa_bg (yehudasa_b@static-66-14-234-139.bdsl.verizon.net) Quit (Ping timeout: 480 seconds)
[4:14] * cattelan (~cattelan@c-66-41-26-220.hsd1.mn.comcast.net) has joined #ceph
[4:23] * yehudasa_bg (~yehudasa_@99-48-179-68.lightspeed.irvnca.sbcglobal.net) has joined #ceph
[4:24] * yehudasa_bg (~yehudasa_@99-48-179-68.lightspeed.irvnca.sbcglobal.net) Quit ()
[4:32] <Qten> Hi All, does anyone know any tips of doco which explains how to use ceph RBD with nova?
[4:35] * Ryan_Lane (~Adium@ Quit (Quit: Leaving.)
[4:38] * sjustlaptop (~sam@24-205-39-1.dhcp.gldl.ca.charter.com) Quit (Quit: Leaving.)
[5:15] * yehuda_hm (~yehuda@99-48-179-68.lightspeed.irvnca.sbcglobal.net) has joined #ceph
[5:19] * jefferai (~quassel@quassel.jefferai.org) Quit (Ping timeout: 480 seconds)
[5:24] * joshd (~jdurgin@ has joined #ceph
[5:27] * jefferai (~quassel@quassel.jefferai.org) has joined #ceph
[5:53] * lofejndif (~lsqavnbok@1RDAABODN.tor-irc.dnsbl.oftc.net) Quit (Quit: gone)
[6:25] * gregaf1 (~Adium@aon.hq.newdream.net) has joined #ceph
[6:26] * gregaf (~Adium@aon.hq.newdream.net) Quit (Read error: Operation timed out)
[6:35] * Ryan_Lane (~Adium@c-98-210-205-93.hsd1.ca.comcast.net) has joined #ceph
[6:54] * bchrisman (~Adium@c-76-103-130-94.hsd1.ca.comcast.net) has joined #ceph
[8:08] * iggy (~iggy@theiggy.com) Quit ()
[8:16] <CristianDM> Hi Qten
[8:16] <CristianDM> I have the same issue. I setup ceph with nova, and now remove them
[8:17] <CristianDM> I can attach corretly volumes, and via this channel I can really get any help
[8:17] <CristianDM> So, without help, ceph don??t work for me
[8:40] * hijacker (~hijacker@ Quit (Remote host closed the connection)
[8:43] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[8:46] * joshd (~jdurgin@ Quit (Quit: Leaving.)
[8:46] * s[X]_ (~sX]@eth589.qld.adsl.internode.on.net) has joined #ceph
[8:53] * s[X] (~sX]@eth589.qld.adsl.internode.on.net) Quit (Ping timeout: 480 seconds)
[9:00] <filoo_absynth> morning shift reporting for duty
[9:03] * verwilst (~verwilst@d5152FEFB.static.telenet.be) has joined #ceph
[9:13] * hijacker (~hijacker@ has joined #ceph
[9:18] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[9:31] * s[X]_ (~sX]@eth589.qld.adsl.internode.on.net) Quit (Remote host closed the connection)
[9:54] * morse (~morse@supercomputing.univpm.it) Quit (Quit: Bye, see you soon)
[9:54] * izdubar (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[9:55] * morse (~morse@supercomputing.univpm.it) has joined #ceph
[9:56] * s[X]_ (~sX]@ppp59-167-157-96.static.internode.on.net) has joined #ceph
[9:59] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[10:09] * CristianDM (~CristianD@host217.190-230-240.telecom.net.ar) Quit (Ping timeout: 480 seconds)
[10:09] * CristianDM (~CristianD@host217.190-230-240.telecom.net.ar) has joined #ceph
[10:28] * LarsFronius (~LarsFroni@testing78.jimdo-server.com) has joined #ceph
[10:36] * adjohn (~adjohn@p3167-ipngn3601marunouchi.tokyo.ocn.ne.jp) has joined #ceph
[10:56] * LarsFronius (~LarsFroni@testing78.jimdo-server.com) Quit (Remote host closed the connection)
[10:57] * LarsFronius (~LarsFroni@testing78.jimdo-server.com) has joined #ceph
[11:00] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[11:01] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Read error: Connection reset by peer)
[11:05] * izdubar (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[11:27] <todin> filoo_absynth: morning
[11:30] * rosco (~r.nap@ Quit (Quit: leaving)
[11:30] * rosco (~r.nap@ has joined #ceph
[11:43] * yoshi (~yoshi@p3167-ipngn3601marunouchi.tokyo.ocn.ne.jp) Quit (Remote host closed the connection)
[11:43] * adjohn (~adjohn@p3167-ipngn3601marunouchi.tokyo.ocn.ne.jp) Quit (Quit: adjohn)
[12:02] * hijacker (~hijacker@ Quit (Quit: Leaving)
[12:06] <S0NiC_> hi guys can anyone help me, i setup ceph 0.45 with fedora 17 ( http://derekweitzel.blogspot.de/2012/02/ceph-on-fedora-16.html) and it works well till i try to mount with the following command mount -t ceph /mnt i got every time libceph mon connection failed but i can ping the machine and firewall is disabled... any hints?
[12:09] <ceph-test> mount -t ceph /mnt/ceph -o name=admin,secret=AkALI5FP4MW5GBAA/EK4lJ635VYFtlk1TjuUtg==
[12:10] <S0NiC_> ceph-test: to me?
[12:10] <ceph-test> yes
[12:10] <S0NiC_> ceph-test: i have to replace the scret with my keyring secret right?
[12:11] <ceph-test> yes
[12:13] <S0NiC_> ceph-test: hmm the output is the same
[12:13] <ceph-test> get output from "ceph -w"
[12:13] <S0NiC_> mount error 5 = input/output error, thats what i got after mount command
[12:15] <S0NiC_> ceph-test: hmm shows nothing... here is my ceph.conf : http://nopaste.info/78181240f9.html
[12:16] <S0NiC_> maybe is there a fault?
[12:17] * hijacker (~hijacker@ has joined #ceph
[12:19] <ceph-test> does ceph-mon,ceph-mds work? Check logs file: /var/log/ceph
[12:22] <S0NiC_> ceph-test: on my testM machine the log mon.a and osd.0 have some content
[12:22] <S0NiC_> i see on the osd.0 log "mount did not detect btrfs"
[12:22] <S0NiC_> i only want to use ext4 for the moment
[12:22] <S0NiC_> maybe thats the fault?
[12:23] <S0NiC_> in my config stands nothing about btrfs
[12:23] <S0NiC_> do i have to create first a filesystem on the /srv/osd.0?
[12:25] <ceph-test> if you use install script it did all things
[12:26] <ceph-test> for osd you need filestore_xattr_use_omap = 1 at config for ext4 (and prepare ext4 for xattr)
[12:26] <S0NiC_> but is this possible the fault that i cant connect to the device?
[12:27] <ceph-test> osd not started without filestore_xattr_use_omap
[12:28] <S0NiC_> ok ceph-test let me first check the kernel if xattr is enabled
[12:28] * hijacker (~hijacker@ Quit (Remote host closed the connection)
[12:29] <S0NiC_> where i have to place the filestore... at the config or doesnt mather?
[12:30] <S0NiC_> in the osd block?
[12:30] <ceph-test> doesnt mather but it was init'ed by install script
[12:31] <S0NiC_> ceph-test: as i told, i only used the howto mentioned above.. maybe i mssed it?
[12:31] <S0NiC_> i
[12:32] <S0NiC_> i didnt use a installsciprt ...
[12:32] <S0NiC_> if i see it right...
[12:33] * rosco (~r.nap@ Quit (Quit: leaving)
[12:33] * rosco (~r.nap@ has joined #ceph
[12:53] * hijacker (~hijacker@ has joined #ceph
[13:05] * verwilst (~verwilst@d5152FEFB.static.telenet.be) Quit (Quit: Ex-Chat)
[13:08] * hijacker (~hijacker@ Quit (Remote host closed the connection)
[13:09] * CristianDM (~CristianD@host217.190-230-240.telecom.net.ar) Quit (Ping timeout: 480 seconds)
[13:09] * CristianDM (~CristianD@host217.190-230-240.telecom.net.ar) has joined #ceph
[13:33] * verwilst (~verwilst@d5152FEFB.static.telenet.be) has joined #ceph
[13:47] <S0NiC_> ceph-test: it works....
[13:49] <S0NiC_> i missed the hostnames.... ;(
[13:49] <S0NiC_> but with whcich fs cephfs writes now?
[13:50] <S0NiC_> because i havent set the filestore_xattr_use_omap = 1 yet
[14:06] * hijacker (~hijacker@ has joined #ceph
[14:12] * hijacker (~hijacker@ Quit (Remote host closed the connection)
[14:12] * hijacker (~hijacker@ has joined #ceph
[14:13] * hijacker (~hijacker@ Quit (Remote host closed the connection)
[14:38] * hijacker (~hijacker@ has joined #ceph
[14:40] * hijacker (~hijacker@ Quit (Remote host closed the connection)
[14:41] * hijacker (~hijacker@ has joined #ceph
[14:48] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) Quit (Ping timeout: 480 seconds)
[15:03] * lofejndif (~lsqavnbok@28IAAEL4K.tor-irc.dnsbl.oftc.net) has joined #ceph
[15:40] * oliver1 (~oliver@p4FFFE57D.dip.t-dialin.net) has joined #ceph
[15:41] * Theuni (~Theuni@ has joined #ceph
[15:56] * cattelan is now known as cattelan_away
[16:01] * nhorman (~nhorman@99-127-245-201.lightspeed.rlghnc.sbcglobal.net) has joined #ceph
[16:09] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[16:19] * s[X]_ (~sX]@ppp59-167-157-96.static.internode.on.net) Quit (Remote host closed the connection)
[16:54] * RNZ (~RNZ@ Quit (Quit: leave)
[16:55] * RNZ (~RNZ@ has joined #ceph
[16:57] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) has joined #ceph
[16:57] * joao (~JL@aon.hq.newdream.net) has joined #ceph
[17:01] * verwilst (~verwilst@d5152FEFB.static.telenet.be) Quit (Quit: Ex-Chat)
[17:07] * iggy (~iggy@theiggy.com) has joined #ceph
[17:13] * nhorman_ (~nhorman@99-127-245-201.lightspeed.rlghnc.sbcglobal.net) has joined #ceph
[17:13] * nhorman_ (~nhorman@99-127-245-201.lightspeed.rlghnc.sbcglobal.net) Quit ()
[17:24] * adjohn (~adjohn@ai126241039157.tik.access-internet.ne.jp) has joined #ceph
[17:24] * BManojlovic (~steki@ Quit (Quit: Ja odoh a vi sta 'ocete...)
[17:24] * adjohn (~adjohn@ai126241039157.tik.access-internet.ne.jp) Quit ()
[17:31] * lofejndif (~lsqavnbok@28IAAEL4K.tor-irc.dnsbl.oftc.net) Quit (Quit: gone)
[17:32] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) Quit (Ping timeout: 480 seconds)
[17:32] * aliguori (~anthony@cpe-70-123-145-39.austin.res.rr.com) Quit (Quit: Ex-Chat)
[17:37] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[17:53] * BManojlovic (~steki@ has joined #ceph
[17:58] * Anticimex (anticimex@netforce.csbnet.se) Quit (Ping timeout: 480 seconds)
[18:00] * aliguori (~anthony@ has joined #ceph
[18:02] * Anticimex (anticimex@netforce.csbnet.se) has joined #ceph
[18:03] * Tv_ (~tv@aon.hq.newdream.net) has joined #ceph
[18:04] * jlogan (~chatzilla@ has joined #ceph
[18:11] * RNZ (~RNZ@ Quit (Quit: leave)
[18:28] <elder> vim
[18:28] <elder> Oops.
[18:31] <Tv_> that reminds me, one thing that happened at the UDS was, i watched a guy intentionally run sudo rm -rf /etc
[18:31] * oliver1 (~oliver@p4FFFE57D.dip.t-dialin.net) has left #ceph
[18:31] <Tv_> that was a good demo of using lxc copy-on-write containers ;)
[18:54] * Anticimex (anticimex@netforce.csbnet.se) Quit (Ping timeout: 480 seconds)
[18:58] * Anticimex (anticimex@netforce.csbnet.se) has joined #ceph
[18:58] * joao (~JL@aon.hq.newdream.net) Quit (Ping timeout: 480 seconds)
[19:04] * joao (~JL@aon.hq.newdream.net) has joined #ceph
[19:06] <Tv_> sagewk: fyi gitbuilder just went red: http://ceph.newdream.net/gitbuilder-oneiric-deb-amd64/#origin/master
[19:06] <sagewk> grr ok
[19:19] * hio (~hio@88-117-119-251.adsl.highway.telekom.at) has joined #ceph
[19:19] * chutzpah (~chutz@ has joined #ceph
[19:34] * jlogan_ (~chatzilla@2600:c00:3010:1:50f7:31fe:9fb:f27c) has joined #ceph
[19:34] * jlogan (~chatzilla@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * aliguori (~anthony@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * rosco (~r.nap@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * elder (~elder@aon.hq.newdream.net) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * nhm (~nh@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * detaos (~quassel@c-50-131-106-101.hsd1.ca.comcast.net) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * ivan\ (~ivan@108-213-76-179.lightspeed.frokca.sbcglobal.net) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * fzylogic (~fzylogic@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * exel (~pi@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * MK_FG (~MK_FG@ Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * guido_ (~guido@mx1.hannover.ccc.de) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * gohko (~gohko@natter.interq.or.jp) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * josef (~seven@nat-pool-rdu.redhat.com) Quit (synthon.oftc.net oxygen.oftc.net)
[19:34] * jlogan_ is now known as jlogan
[19:35] * jlogan (~chatzilla@2600:c00:3010:1:50f7:31fe:9fb:f27c) Quit (Killed (synthon.oftc.net (Nick collision (new))))
[19:35] * jlogan (~chatzilla@ has joined #ceph
[19:35] * aliguori (~anthony@ has joined #ceph
[19:35] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) has joined #ceph
[19:35] * rosco (~r.nap@ has joined #ceph
[19:35] * elder (~elder@aon.hq.newdream.net) has joined #ceph
[19:35] * nhm (~nh@ has joined #ceph
[19:35] * detaos (~quassel@c-50-131-106-101.hsd1.ca.comcast.net) has joined #ceph
[19:35] * ivan\ (~ivan@108-213-76-179.lightspeed.frokca.sbcglobal.net) has joined #ceph
[19:35] * fzylogic (~fzylogic@ has joined #ceph
[19:35] * josef (~seven@nat-pool-rdu.redhat.com) has joined #ceph
[19:35] * exel (~pi@ has joined #ceph
[19:35] * MK_FG (~MK_FG@ has joined #ceph
[19:35] * gohko (~gohko@natter.interq.or.jp) has joined #ceph
[19:35] * guido_ (~guido@mx1.hannover.ccc.de) has joined #ceph
[19:35] * jlogan_ (~chatzilla@2600:c00:3010:1:50f7:31fe:9fb:f27c) has joined #ceph
[19:40] * jlogan (~chatzilla@ Quit (Ping timeout: 480 seconds)
[19:40] * jlogan_ is now known as jlogan
[19:41] * izdubar (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[19:47] <elder> Tv_ what do you mean "went red"?
[19:48] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[19:48] <elder> And how does one determine the build machines are fucntioning? I know my push from this morning didn't kick off a build.
[19:48] <elder> I don't know how to get that fixed when that happens.
[19:54] * hio (~hio@88-117-119-251.adsl.highway.telekom.at) Quit (Quit: Leaving)
[19:56] <Tv_> elder: the gitbuilders are in a state of mess, sadly
[19:56] <elder> OK.
[19:56] <Tv_> elder: i'll fix them (/work with our new admin to get them fixed) once the osd hotplugging work settles down
[19:56] <Tv_> elder: but if you follow the link i pasted, you see state of master
[19:56] <Tv_> for that one particular gitbuilder
[19:56] <Tv_> though now it's back to normal
[19:57] <elder> I'm not desparate for it now. At this point I'm just doing a sort of validation build because I haven't done it in a week or two.
[19:57] <Tv_> see e.g. wip-mon-noname on that page for an example of where a commit made an error appear
[19:58] <elder> OK, I get the "commit created a bad build." My issue was that I wasn't getting a build at all.
[19:58] <Tv_> elder: that's.. not really possible ;)
[19:58] <Tv_> elder: maybe it was slow. point me to specifics for troubleshooting
[20:00] <elder> ceph-client Actually, now that you asked I checked and it turns out my branch isn't updated on the git tree for some reason despite my push.
[20:00] <elder> Hmm. Must have missed something.
[20:00] <elder> Just a minute.
[20:00] * izdubar (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Quit: Leaving)
[20:02] * hio (~hio@88-117-119-251.adsl.highway.telekom.at) has joined #ceph
[20:04] <elder> I must have done something stupid I pushed my branch "again" and it's good now.
[20:05] <Tv_> elder: my bet is on "didn't notice the error message you got the first time"
[20:05] <elder> Well, I noticed, and I *thought* I re-did it with a -f flag. the error was non-fast-forward branch.
[20:05] <elder> But yes, that general thing is my bet as well.
[20:05] * Tv_ dislikes -f
[20:05] <Tv_> too easy to "git push -f" and do a world of hurt
[20:06] <Tv_> as in, push all the branches
[20:06] <Tv_> i prefer "git push myremote +mybranch" for forcing
[20:06] <Tv_> but then again, i always push just individual branches, for much the same reason
[20:06] <elder> I hadn't thought of using that syntax and didn't realize I could but it fits with the fetch syntax.
[20:07] <elder> I do too.
[20:07] <elder> I'll switch to using the + to indicate force.
[20:07] <Tv_> my mnemonic is "i'm positive about this" ;)
[20:08] <elder> More of a thinkmonic it seems
[20:09] <elder> But now (of course) I looked it up and "mnemonic" is probably right.
[20:21] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) has joined #ceph
[20:23] <hio> not quite sure about libtool stuff, but I tried to compile from source just now and it only warned me mid-make compilation that curl and libexpat are needed as well
[20:33] <Tv_> hio: there is an explicit check for curl in the top-level autoconf
[20:35] * jlogan (~chatzilla@2600:c00:3010:1:50f7:31fe:9fb:f27c) Quit (Quit: ChatZilla [Firefox 13.0/20120501201020])
[20:36] * Ryan_Lane (~Adium@c-98-210-205-93.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[20:41] <Tv_> hio: if you can provide a pastebin of the whole log (starting from ./configure, showing commands you ran), i'll happily dig in more
[20:48] * LarsFronius (~LarsFroni@testing78.jimdo-server.com) Quit (Quit: LarsFronius)
[20:59] * aa (~aa@r200-40-114-26.ae-static.anteldata.net.uy) has joined #ceph
[21:01] <hio> Tv_, I reconfigured each time I installed another missing library, so the log is clean. I managed to compile it now, I just thought I'd mention it, maybe someone forgot to add curl to the dependencies
[21:13] * andreask (~andreas@chello062178013131.5.11.vie.surfer.at) has joined #ceph
[21:23] <CristianDM> Hi. Any tutorial to work with openstack and boot the vm from ceph?
[21:23] <CristianDM> Currently I can boot VM from ceph and have inestability working with ceph
[21:41] * lofejndif (~lsqavnbok@82VAADQYY.tor-irc.dnsbl.oftc.net) has joined #ceph
[21:54] <Tv_> CristianDM: unstability?
[21:55] <Tv_> CristianDM: I don't think we've gotten around to writing documentation on the OpenStack integration; if you're having problems, please tell us more about them
[21:57] <Tv_> elder: you might want to look at the email thread with subject "Ceph kernel client - kernel crashes"
[21:57] <CristianDM> I try to get help for three days now
[21:57] <CristianDM> In this moment I unninstall ceph and I checking with swift
[21:58] <Tv_> elder: last message, past the point where he disables NX, has a paging failure that apparently only occurs with ceph kclient, and then there's an unrelated what looks like easy-to-reproduce ceph.ko bug
[21:58] <CristianDM> Good software but without help, is the same as nothing.
[21:59] <CristianDM> Docs are broken links and any help for the most biggest utility of ceph
[21:59] <CristianDM> Openstack..
[21:59] <Tv_> CristianDM: please email the list about any broken links you still see in the docs
[22:00] <CristianDM> Ok, but I try for three days get any help in the IRC
[22:00] <CristianDM> If ceph will have IRC channel, this is for any help
[22:00] <Tv_> CristianDM: if you want guarantees of support, i can point you to the right people who sell commercial support
[22:01] <Tv_> CristianDM: if you want to use the open source project as-is, you have to accommodate for the fact that all of the team traveled to Vegas earlier this week
[22:01] <CristianDM> Not guarantees of support
[22:01] <CristianDM> The said "Open source, free, bla bla bla"
[22:01] <CristianDM> So, if ceph will help the comunity, the minimum is a documentation about how to use it
[22:01] <CristianDM> I can view any tutorial, how-to about ceph and openstack
[22:01] <darkfader> i think he doesn't care your team was away and had no chance of reading his questions.
[22:02] <darkfader> like, because.
[22:02] <Tv_> care goes both ways
[22:03] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) Quit (Quit: Zzzzzz)
[22:04] <Tv_> we also hired a full-time technical writer
[22:04] <Tv_> and the recent doc revamp you've seen was his first chunk of work
[22:04] <darkfader> Tv_: well i'm fully convinced your doing the right work on the docs
[22:05] <CristianDM> Tv_ for example, where I can fin??t docs about openstack and ceph?
[22:05] <darkfader> i don't think there's any sense in doing an openstack howto before the core ceph docs are in shape
[22:05] <Tv_> darkfader: i'm personally psyched about the new osd hotplugging stuff, and how much simpler install/management becomes
[22:05] <CristianDM> yes? So, how boot a VM form ceph?
[22:06] <darkfader> CristianDM: just *maybe* you'll have to be the one providing docs
[22:06] <darkfader> spoon-feed etc etc i'm out
[22:07] <CristianDM> I can??t, I don??t know how work ceph. I am user of openstack
[22:07] <CristianDM> If I said that X software can run x operation but without docs
[22:08] <CristianDM> So, how run the operation?
[22:09] <Tv_> CristianDM: OpenStack doesn't support cloning Glance images into Nova volumes. Setting up Ceph RBD as the block device for vms is not a 5-minute tutorial at this point.
[22:10] <CristianDM> Tv_ Not need 5 minute tutorial.
[22:11] <CristianDM> Tv_: But any help about the setup. I setup ceph with openstack, but I can??t boot VM, when try to attach a volume created this don??t attach.
[22:13] <iggy> it's maybe a bit early to be pushing into that territory unless you're a dev or follow both projects pretty closely
[22:13] <CristianDM> Tv_: Why don??t have any page in docs that doc the features with ceph at today, and the roadmap
[22:15] <Tv_> CristianDM: I'm not sure how to help you at this point. Not everything has been distilled into a document. You either need to be prepared to learn, or you need to be prepared to get commercial support.
[22:15] * joshd (~joshd@aon.hq.newdream.net) has joined #ceph
[22:15] <CristianDM> Tv_: I am prepared to read, but I don??t know where read
[22:16] <Tv_> CristianDM: Mailing lists. Source code. Google.com.
[22:17] <darkfader> i think /ig scales best
[22:17] <CristianDM> Tv_: Yes, and I can find any how-to setup openstack with ceph. Really, I use google for days and try to get a working system
[22:17] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) has joined #ceph
[22:18] <CristianDM> Tv_: But I only get a creating volumes, remove from ceph, but never I can attach this to VM.
[22:18] <Tv_> CristianDM: As I said earlier, not everything has been distilled into a document.
[22:18] <Tv_> CristianDM: I did not imply mailing lists, source code, or google would lead you to one.
[22:18] <CristianDM> Tv_: No document, but in chats or in the mailing lists
[22:18] <Tv_> CristianDM: I meant to imply they will lead you to the raw information.
[22:19] * MarkDude (~MT@c-71-198-138-155.hsd1.ca.comcast.net) Quit (Read error: Connection reset by peer)
[22:21] <CristianDM> Tv_: So how the users will adopt ceph? Without info about how-to use it?
[22:21] <CristianDM> Tv_: The only idea is get commercial support and cash?
[22:21] * nhorman (~nhorman@99-127-245-201.lightspeed.rlghnc.sbcglobal.net) Quit (Quit: Leaving)
[22:22] <ajm> or simply be willing to learn, and not necessarily from a "howto" document that spells out all the steps for you.
[22:22] <CristianDM> I read users, mailing lists, etc
[22:22] <morpheus> we set up our cluster in some weeks of try and error, reading lots of old irc logs, mailing list posts and so on
[22:22] <Tv_> CristianDM: documentation is improving literally as we speak, I'm not sure what more I can do for you
[22:22] <ajm> you'd probably have to also go back to the source of one or both projects occasionally to figure that out.
[22:23] * Oliver1 (~oliver1@p54839588.dip.t-dialin.net) has joined #ceph
[22:23] <ajm> with something as in-flux as both projects are, sometimes it doesn't make sense to document things when you know they are going to change.
[22:23] <Tv_> CristianDM: most of the real deployments of ceph will be on hundreds of servers, with capable admins who've already picked up a lot of know-how
[22:23] <ajm> morpheus: me as well :)
[22:24] <Tv_> CristianDM: at the scale where you're building custom billing solutions from scratch, weeks of work on integrating your core software is not a significant barrier
[22:25] <ajm> "weeks" or a case of redbull and a long weekend :)
[22:25] * Oliver1 (~oliver1@p54839588.dip.t-dialin.net) Quit ()
[22:25] <Tv_> on the scale this thing is, most admins insist on months of burn-in
[22:25] <CristianDM> Tv_: Weeks to mount a rbd volume?
[22:25] <ajm> oh yes, i just mean to learn how to get it working
[22:25] <iggy> my old boss used to provide red bull if i worked late
[22:26] <Tv_> CristianDM: perhaps we are having a language barrier issue here
[22:26] <CristianDM> Tv_: I don??t understand your idea.
[22:26] <darkfader> mine is in the progress of purchasing floor lamp just as of now
[22:26] <ajm> CristianDM: if your starting from 0 knowledge, it could take you a week to figure everything out and get your first success, yes.
[22:26] <darkfader> 3 weeks after our move
[22:26] <iggy> it's easy as pie to simply mount an rbd volume... the hard part is integrating 8 different pieces of software to use it
[22:27] <Tv_> i wish pie was that easy, i'd eat pie more often
[22:27] <morpheus> :)
[22:27] * Ryan_Lane (~Adium@ has joined #ceph
[22:28] <ajm> Tv_: pie is easy!
[22:28] <Tv_> ajm: you haven't seen me bake
[22:28] <Tv_> ajm: just cleaning up afterwards will take more time
[22:28] <hio> I just had a "great" idea: Is it terribly stupid to use ceph as a key-value database? I mean, it writes objects, right?
[22:28] <ajm> lol
[22:28] <Tv_> hio: each object has a key-value store on the side, those are ultimately stored in leveldb
[22:29] <hio> Tv_, yes I read about that, inside the xattr
[22:29] <Tv_> hio: as long as you understand that ceph is a full-consistency model and not eventual consistency, it might work just fine
[22:29] <CristianDM> Tv_: I only need to use ceph with openstack, and the logig here not are 8 pieces of software, maybe 4
[22:29] <Tv_> hio: not xattr, look up "omap"
[22:30] <hio> Tv_, Okay, thanks, I'll look it up. Well, full consistency is basically better, I do the caching with memcached myself then
[22:30] <Tv_> CristianDM: i'm at 7 in my head right now... kvm qemu libvirt nova ceph-mon ceph-osd
[22:30] <Tv_> hio: just different performance implications than e.g. Cassandra
[22:30] <CristianDM> Tv_: And when I try to get help, the help is read google?
[22:31] <CristianDM> Tv_: ceph-mon, osd, etc not is the issue
[22:31] <Tv_> CristianDM: you kept asking for a document that will be ready in maybe 4 weeks
[22:31] <CristianDM> Tv_: Build the ceph cluster is very easy
[22:31] <CristianDM> Tv_: The issue is when try to use this as rbd
[22:32] <Tv_> CristianDM: not to mention i still don't know whether you really understood the limitations of OpenStack Essex as it comes to volumes
[22:32] <hio> Tv_, I'm not very versed with filesystems yet, so as far as I understand, I will need to try and do as few calls to ceph as possible because it fsyncs every single time.
[22:32] <Tv_> hio: not quite but in that general direction
[22:33] <Tv_> hio: a single operation is costlier to a full-consistency system than to an eventual-consistency system.. or "can be" costlier, depending on a bunch of things
[22:34] <hio> But since the metadata is stored in leveldb, it should actually be comparable to other key-value stores, no?
[22:34] <Tv_> hio: the full-consistency part means that Ceph needs to jump through a few extra hoops, worry about journals and replication and what not
[22:35] <CristianDM> Tv_: I setup nova to use ceph corretly, but then try to attach a volume into a intance I can??t.
[22:35] <Tv_> hio: and the typical workloads we see are not the "small values" workloads typical key-value stores see, so ceph is probably not very optimized for that use case
[22:37] <iggy> CristianDM: sounds like an openstack problem, not a ceph problem
[22:38] <CristianDM> iggy: Or of the ceph connector
[22:38] <iggy> which was written by whom
[22:39] <CristianDM> Who is whom?
[22:40] <hio> Hm well, I'll just have to try it I suppose. I like the idea of forgetting about databases, it's like removing an entire dimension of additional administrative overhead
[22:44] * Ryan_Lane (~Adium@ Quit (Quit: Leaving.)
[22:49] * Theuni (~Theuni@ Quit (Quit: Leaving.)
[22:51] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) Quit (Read error: Connection reset by peer)
[22:53] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) has joined #ceph
[22:53] * Ryan_Lane (~Adium@ has joined #ceph
[22:55] * joshd (~joshd@aon.hq.newdream.net) Quit (Ping timeout: 480 seconds)
[22:55] * Ryan_Lane (~Adium@ Quit ()
[23:05] * joshd (~joshd@p4FFFE57D.dip.t-dialin.net) has joined #ceph
[23:09] <joshd> CristianDM: are there errors in any of the nova logs when the attach fails?
[23:16] <CristianDM> joshd: Hi. Thanks for your help
[23:17] <CristianDM> joshd: No, I am checked all the logs but wihout any error.
[23:17] <CristianDM> joshd: This show "Attaching..." and next "Active"
[23:18] * elder (~elder@aon.hq.newdream.net) Quit (Remote host closed the connection)
[23:21] * gregaf (~Adium@aon.hq.newdream.net) has joined #ceph
[23:22] * elder (~elder@aon.hq.newdream.net) has joined #ceph
[23:23] * joshd (~joshd@p4FFFE57D.dip.t-dialin.net) Quit (Ping timeout: 480 seconds)
[23:24] * gregaf1 (~Adium@aon.hq.newdream.net) Quit (Ping timeout: 480 seconds)
[23:28] * nyeates (~nyeates@pool-108-40-29-119.bltmmd.fios.verizon.net) Quit (Quit: Zzzzzz)
[23:38] * joshd (~joshd@aon.hq.newdream.net) has joined #ceph
[23:44] * hio (~hio@88-117-119-251.adsl.highway.telekom.at) Quit (Quit: Leaving)
[23:47] * dennisj_ (~chatzilla@p5DCF7BF7.dip.t-dialin.net) has joined #ceph
[23:48] <joshd> CristianDM: what symptoms are you seeing of attach failing then?
[23:52] * Ryan_Lane (~Adium@ has joined #ceph
[23:53] * dennisj (~chatzilla@p5DCF6885.dip.t-dialin.net) Quit (Ping timeout: 480 seconds)
[23:59] * joshd (~joshd@aon.hq.newdream.net) Quit (Ping timeout: 480 seconds)
[23:59] * s[X]_ (~sX]@eth589.qld.adsl.internode.on.net) has joined #ceph

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