#ceph IRC Log

Index

IRC Log for 2014-01-03

Timestamps are in GMT/BST.

[0:00] * dalgaaf (uid15138@id-15138.ealing.irccloud.com) Quit (Read error: Connection reset by peer)
[0:00] * dalgaaf (uid15138@id-15138.ealing.irccloud.com) has joined #ceph
[0:02] * alexp789 (~alexp789@host81-151-154-67.range81-151.btcentralplus.com) Quit (Remote host closed the connection)
[0:03] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) Quit (Remote host closed the connection)
[0:04] * jerker_ (jerker@82ee1319.test.dnsbl.oftc.net) Quit (Read error: Connection reset by peer)
[0:04] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) has joined #ceph
[0:08] * jerker (jerker@Psilocybe.Update.UU.SE) has joined #ceph
[0:12] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) Quit (Ping timeout: 480 seconds)
[0:14] * fireD (~fireD@93-139-180-225.adsl.net.t-com.hr) has joined #ceph
[0:24] * zerick (~eocrospom@190.187.21.53) Quit (Read error: Connection reset by peer)
[0:25] * _zerick_ (~eocrospom@190.187.21.53) has joined #ceph
[0:26] * _zerick_ (~eocrospom@190.187.21.53) Quit (Read error: Connection reset by peer)
[0:28] * _zerick_ (~eocrospom@190.187.21.53) has joined #ceph
[0:28] * diegows (~diegows@190.190.17.57) has joined #ceph
[0:36] * dmsimard (~Adium@69-165-206-93.cable.teksavvy.com) has joined #ceph
[0:46] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) has joined #ceph
[0:47] * sarob_ (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) has joined #ceph
[0:47] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) Quit (Read error: Connection reset by peer)
[0:47] * _zerick_ (~eocrospom@190.187.21.53) Quit (Quit: Saliendo)
[0:47] * zerick (~eocrospom@190.187.21.53) has joined #ceph
[1:00] * iaXe (axe@123.116.43.193) Quit (Ping timeout: 480 seconds)
[1:09] * mattbenjamin (~matt@aa2.linuxbox.com) Quit (Quit: Leaving.)
[1:09] * bjornar (~bjornar@ti0099a340-dhcp0395.bb.online.no) has joined #ceph
[1:10] * dmsimard (~Adium@69-165-206-93.cable.teksavvy.com) Quit (Quit: Leaving.)
[1:13] * seif (sid11725@2604:8300:100:200b:6667:2:0:2dcd) Quit (Read error: Connection reset by peer)
[1:14] * seif (sid11725@id-11725.ealing.irccloud.com) has joined #ceph
[1:25] * scuttlemonkey_ (~scuttlemo@c-69-244-181-5.hsd1.mi.comcast.net) has joined #ceph
[1:28] * KevinPerks (~Adium@cpe-066-026-252-218.triad.res.rr.com) Quit (Quit: Leaving.)
[1:31] * scuttlemonkey (~scuttlemo@c-69-244-181-5.hsd1.mi.comcast.net) Quit (Ping timeout: 480 seconds)
[1:37] * ircolle (~Adium@2601:1:8380:2d9:b8f9:3861:a64d:6238) Quit (Quit: Leaving.)
[1:38] * dalgaaf (uid15138@id-15138.ealing.irccloud.com) Quit (Ping timeout: 480 seconds)
[1:39] * dalgaaf (~uid15138@id-15138.ealing.irccloud.com) has joined #ceph
[1:44] * yanzheng (~zhyan@jfdmzpr03-ext.jf.intel.com) has joined #ceph
[1:45] * KevinPerks (~Adium@cpe-066-026-252-218.triad.res.rr.com) has joined #ceph
[1:46] * AfC (~andrew@jim1020952.lnk.telstra.net) has joined #ceph
[1:52] * mozg (~andrei@46.229.149.194) Quit (Ping timeout: 480 seconds)
[1:54] * LeaChim (~LeaChim@host86-161-89-52.range86-161.btcentralplus.com) Quit (Ping timeout: 480 seconds)
[2:01] * mschiff_ (~mschiff@port-6279.pppoe.wtnet.de) has joined #ceph
[2:01] * xarses (~andreww@12.164.168.115) Quit (Read error: Connection reset by peer)
[2:01] * xarses (~andreww@12.164.168.115) has joined #ceph
[2:02] * mattbenjamin (~matt@76-206-42-105.lightspeed.livnmi.sbcglobal.net) has joined #ceph
[2:05] * bjornar (~bjornar@ti0099a340-dhcp0395.bb.online.no) Quit (Ping timeout: 480 seconds)
[2:08] * zerick (~eocrospom@190.187.21.53) Quit (Remote host closed the connection)
[2:09] * mschiff (~mschiff@port-2825.pppoe.wtnet.de) Quit (Ping timeout: 480 seconds)
[2:11] * Cube (~Cube@66-87-65-110.pools.spcsdns.net) Quit (Quit: Leaving.)
[2:14] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[2:25] * xarses (~andreww@12.164.168.115) Quit (Ping timeout: 480 seconds)
[2:29] * Tamil1 (~Adium@cpe-76-168-18-224.socal.res.rr.com) Quit (Quit: Leaving.)
[2:30] * ScOut3R (~scout3r@540205B4.dsl.pool.telekom.hu) Quit ()
[2:32] * angdraug (~angdraug@12.164.168.115) Quit (Quit: Leaving)
[2:39] * AfC (~andrew@jim1020952.lnk.telstra.net) Quit (Quit: Leaving.)
[2:39] * Tamil1 (~Adium@cpe-76-168-18-224.socal.res.rr.com) has joined #ceph
[2:40] * iaXe (~axe@223.223.202.194) has joined #ceph
[2:47] * twx (~twx@rosamoln.org) has joined #ceph
[2:48] * Karcaw (~evan@68-186-68-219.dhcp.knwc.wa.charter.com) has joined #ceph
[2:48] * DLange is now known as Guest2237
[2:48] * DLange (~DLange@dlange.user.oftc.net) has joined #ceph
[2:48] * acaos_ (~zac@209.99.103.42) has joined #ceph
[2:49] * nolan_ (~nolan@2001:470:1:41:20c:29ff:fe9a:60be) has joined #ceph
[2:49] * twx_ (~twx@rosamoln.org) Quit (Ping timeout: 480 seconds)
[2:49] * Karcaw_ (~evan@68-186-68-219.dhcp.knwc.wa.charter.com) Quit (Ping timeout: 480 seconds)
[2:50] * sig_wall (~adjkru@185.14.185.91) Quit (Ping timeout: 480 seconds)
[2:50] * Jahkeup (Jahkeup@gemini.ca.us.panicbnc.net) Quit (Remote host closed the connection)
[2:50] * Guest2237 (~DLange@dlange.user.oftc.net) Quit (Ping timeout: 480 seconds)
[2:50] * acaos (~zac@209-99-103-42.fwd.datafoundry.com) Quit (Ping timeout: 480 seconds)
[2:50] * blahnana (~bman@blahnana.com) Quit (Ping timeout: 480 seconds)
[2:50] * nolan (~nolan@2001:470:1:41:20c:29ff:fe9a:60be) Quit (Ping timeout: 480 seconds)
[2:50] * nolan_ is now known as nolan
[2:51] * KarlV (~kvietmei@pool-71-160-136-179.lsanca.dsl-w.verizon.net) has joined #ceph
[2:51] * blahnana (~bman@blahnana.com) has joined #ceph
[2:51] * KarlV (~kvietmei@pool-71-160-136-179.lsanca.dsl-w.verizon.net) Quit ()
[2:52] * sig_wall (~adjkru@185.14.185.91) has joined #ceph
[2:52] * wer (~wer@206-248-239-142.unassigned.ntelos.net) Quit (Read error: Connection reset by peer)
[2:53] * wer (~wer@206-248-239-142.unassigned.ntelos.net) has joined #ceph
[2:53] * sarob (~sarob@2001:4998:effd:600:2d10:d367:997:3861) has joined #ceph
[2:53] * sarob (~sarob@2001:4998:effd:600:2d10:d367:997:3861) Quit (Remote host closed the connection)
[2:53] * pm0d (~guest@rdns.pw) has joined #ceph
[2:53] * sarob (~sarob@2001:4998:effd:600:2d10:d367:997:3861) has joined #ceph
[2:54] * machete (~guest@rdns.pw) Quit (Remote host closed the connection)
[2:55] * IndraPr (~indra@bb121-6-34-250.singnet.com.sg) Quit ()
[2:57] * Jahkeup (~Jahkeup@gemini.ca.us.panicbnc.net) has joined #ceph
[2:57] * sarob_ (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) Quit (Ping timeout: 480 seconds)
[2:58] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) Quit (Ping timeout: 480 seconds)
[2:59] * KevinPerks (~Adium@cpe-066-026-252-218.triad.res.rr.com) Quit (Quit: Leaving.)
[2:59] * xarses (~andreww@c-24-23-183-44.hsd1.ca.comcast.net) has joined #ceph
[2:59] * beardo (~sma310@beardo.cc.lehigh.edu) Quit (Ping timeout: 480 seconds)
[2:59] * beardo (~sma310@beardo.cc.lehigh.edu) has joined #ceph
[2:59] * mschiff (~mschiff@port-19476.pppoe.wtnet.de) has joined #ceph
[3:01] * sarob (~sarob@2001:4998:effd:600:2d10:d367:997:3861) Quit (Ping timeout: 480 seconds)
[3:07] * mschiff_ (~mschiff@port-6279.pppoe.wtnet.de) Quit (Ping timeout: 480 seconds)
[3:08] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) Quit (Ping timeout: 480 seconds)
[3:13] * mikedawson (~chatzilla@23-25-46-97-static.hfc.comcastbusiness.net) Quit (Ping timeout: 480 seconds)
[3:19] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) has joined #ceph
[3:20] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[3:20] * diegows (~diegows@190.190.17.57) Quit (Ping timeout: 480 seconds)
[3:20] * julian (~julianwa@125.70.133.91) Quit (Quit: afk)
[3:20] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) has joined #ceph
[3:21] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) Quit ()
[3:21] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) has joined #ceph
[3:23] * shang (~ShangWu@175.41.48.77) has joined #ceph
[3:24] * Discovery (~Discovery@192.162.100.197) Quit (Read error: Connection reset by peer)
[3:29] * KevinPerks (~Adium@cpe-066-026-252-218.triad.res.rr.com) has joined #ceph
[3:33] * Machske (~Bram@d5152D87C.static.telenet.be) has joined #ceph
[3:36] * Macheske (~Bram@d5152D87C.static.telenet.be) Quit (Ping timeout: 480 seconds)
[3:36] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[3:37] * ScOut3R (~scout3r@540205B4.dsl.pool.telekom.hu) has joined #ceph
[3:39] * ScOut3R (~scout3r@540205B4.dsl.pool.telekom.hu) Quit ()
[3:40] * KevinPerks (~Adium@cpe-066-026-252-218.triad.res.rr.com) Quit (Ping timeout: 480 seconds)
[3:42] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) has joined #ceph
[3:42] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[3:53] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) Quit (Ping timeout: 480 seconds)
[3:55] * i_m (~ivan.miro@95.180.8.206) Quit (Ping timeout: 480 seconds)
[3:57] * dpippenger (~riven@66-192-9-78.static.twtelecom.net) Quit (Quit: Leaving.)
[4:10] * KindTwo (KindOne@h191.38.28.71.dynamic.ip.windstream.net) has joined #ceph
[4:11] * mkoderer_ (uid11949@id-11949.ealing.irccloud.com) has joined #ceph
[4:11] * KindOne (KindOne@0001a7db.user.oftc.net) Quit (Ping timeout: 480 seconds)
[4:11] * KindTwo is now known as KindOne
[4:13] * mkoderer (uid11949@ealing.irccloud.com) Quit (Read error: Connection reset by peer)
[4:13] * mkoderer_ is now known as mkoderer
[4:16] * Sysadmin88 (~IceChat77@2.218.8.40) Quit (Ping timeout: 480 seconds)
[4:17] * jefferai (~quassel@corkblock.jefferai.org) has joined #ceph
[4:18] * Tamil1 (~Adium@cpe-76-168-18-224.socal.res.rr.com) has left #ceph
[4:20] * Machske (~Bram@d5152D87C.static.telenet.be) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * blahnana (~bman@blahnana.com) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * iaXe (~axe@223.223.202.194) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * jesus (~jesus@emp048-51.eduroam.uu.se) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * jdmason (~jon@192.55.54.40) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * houkouonchi-home (~linux@66-215-209-207.dhcp.rvsd.ca.charter.com) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * Vacum (~vovo@88.130.211.69) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * _matt (~matt@ccpc-mwr.bath.ac.uk) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * athrift_ (~nz_monkey@203.86.205.13) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * paradon (~thomas@60.234.66.253) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * grifferz (~andy@bitfolk.com) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * stj (~s@tully.csail.mit.edu) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * chris_lu (~ccc2@128.180.82.4) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * ferai (~quassel@corkblock.jefferai.org) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * Anticimex (anticimex@95.80.32.80) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * toabctl (~toabctl@toabctl.de) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:20] * ZyTer_ (~ZyTer@ghostbusters.apinnet.fr) Quit (reticulum.oftc.net kinetic.oftc.net)
[4:21] * Xedecimal (~Xedecimal@c-24-22-44-186.hsd1.or.comcast.net) has joined #ceph
[4:21] * Machske (~Bram@d5152D87C.static.telenet.be) has joined #ceph
[4:21] * blahnana (~bman@blahnana.com) has joined #ceph
[4:21] * iaXe (~axe@223.223.202.194) has joined #ceph
[4:21] * jesus (~jesus@emp048-51.eduroam.uu.se) has joined #ceph
[4:21] * jdmason (~jon@192.55.54.40) has joined #ceph
[4:21] * houkouonchi-home (~linux@66-215-209-207.dhcp.rvsd.ca.charter.com) has joined #ceph
[4:21] * Vacum (~vovo@88.130.211.69) has joined #ceph
[4:21] * toabctl (~toabctl@toabctl.de) has joined #ceph
[4:21] * chris_lu (~ccc2@128.180.82.4) has joined #ceph
[4:21] * stj (~s@tully.csail.mit.edu) has joined #ceph
[4:21] * grifferz (~andy@bitfolk.com) has joined #ceph
[4:21] * paradon (~thomas@60.234.66.253) has joined #ceph
[4:21] * Anticimex (anticimex@95.80.32.80) has joined #ceph
[4:21] * athrift_ (~nz_monkey@203.86.205.13) has joined #ceph
[4:21] * _matt (~matt@ccpc-mwr.bath.ac.uk) has joined #ceph
[4:21] * ZyTer_ (~ZyTer@ghostbusters.apinnet.fr) has joined #ceph
[4:22] <Xedecimal> I'm trying to wrap my head around Ceph, would this solution still be good for me if I have a single server loaded up with a series of drives that I wish to combine into one lump sum of storage space? Possibly adding other servers in the future? I've been going over the quick start but it really suggests to me that I want to create 3 nodes, which looks to me as 3 separate servers
[4:23] <janos> Xedecimal: yes ideally
[4:23] <janos> they default config is expecting multiple physical failure domains
[4:23] <janos> ie: one entire machine dies, you're degraded but safe
[4:24] <janos> where that machine is an OSD host which houses half of your harddrives (OSD's)
[4:24] <Xedecimal> but I can run it effectively and should spend the time learning more about it even if I only have 1 server ? I'm mainly worried about individual drives going out in the machine, I can always replace the server hardware and hopefully if the data is redundant that should be safe
[4:24] <janos> you can, but you'd need to edit the crushmap
[4:24] <janos> it's not really made for that
[4:24] <Xedecimal> I was going to say, is this overkill ?
[4:25] <janos> that's mostly seen just for testing
[4:25] <janos> yeah, it is, i'd say
[4:25] <janos> i love ceph, but that scenario is pushing it. it's certainly possible, but you're not really using cepg for ceph ;)
[4:25] <Xedecimal> I've been screwing around with greyhole but the way it's laid out seems ineffective, I ran some tests and destroyed some data and had to run recovery commands to get it back, not to mention it didn't seem to adapt to if I were to remove a drive, and I was reading about systems that can put load on the lowest temperature drives and such
[4:26] <Xedecimal> greyhole seems a bit too simple, ceph seems a bit too advanced, lol
[4:26] <janos> your initial setup sounds likea job for raid
[4:26] <janos> though expanding raid is far less fun
[4:27] <Xedecimal> are there other solutions out there? It seems like just recently distributed file systems are really taking off. I want to be able to yank a bad drive out and the system to just adapt, or add another drive and just forget it, instead of sitting here for hours configuring everything constantly
[4:27] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) has joined #ceph
[4:27] <janos> you could do a bare minimal of 2 OSD hosts and a Mon
[4:27] <janos> though i'd recommend 3 mons
[4:28] <janos> i think ceph is a bit much for one machine
[4:28] <Xedecimal> would both OSD hosts run inside the same machine? Or are OSD hosts individual machines ?
[4:28] <janos> possible, but not recommended
[4:28] <janos> sorry - whe i say osd host i mean physcial machines each with X number of OSD's (usually harrdrives)
[4:29] <Xedecimal> I look up OSD and find open side diagonal seam, lol
[4:29] <Xedecimal> on screen display ?
[4:30] <dmick> object storage daemon
[4:30] <Xedecimal> ahhhhhhhhh!
[4:31] <janos> ah sorry, i stepped away a second
[4:31] <janos> each OSd usually maps to an individual drive. it doesn't have to, but i'd recommend it
[4:32] <Xedecimal> so if I have like 6 drives we'd be expecting to run each drive in it's own machine ?
[4:32] <janos> naw
[4:32] <janos> a machine can host mutiple drives/osds
[4:32] <janos> but if you have 6 drives, and 2 machines - 3 drives per machine
[4:33] <janos> that way if a machine dies, your data is good
[4:33] <dmick> you could certainly one run machine, one osd per drive
[4:33] <dmick> it's no *less* reliable than RAID
[4:33] <Xedecimal> there's not much for alternative solutions though is there ?
[4:33] <janos> right
[4:33] <dmick> zfs or software raid
[4:33] <Xedecimal> raid is a major pain, I don't know how much time I've spent trying to reconfigure this stuff when switching a drive or something
[4:34] <Xedecimal> also, we've got some extra backup drives that we back our data up to from time to time... Why not just use them in this system and benefit from the extra space ?
[4:34] <janos> sorry i have to roll out
[4:35] <Xedecimal> alrighty
[4:41] * nerdtron (~oftc-webi@202.60.8.250) has joined #ceph
[4:41] <nerdtron> hi all
[4:41] <nerdtron> i have a 1 pgs stuck unclean
[4:41] <nerdtron> how do i make my cluster ceph health OK again?
[4:42] <nerdtron> the pg is 2.3f
[4:44] * mschiff (~mschiff@port-19476.pppoe.wtnet.de) Quit (Remote host closed the connection)
[4:44] <dmick> what does ceph pg 2.3f query show about it (don't paste it all here)
[4:48] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) has joined #ceph
[4:48] <nerdtron> dmick: http://pastebin.com/YGFX0DPd
[4:48] <nerdtron> all i see is wrong node? how's that?
[4:49] <dmick> so, that didn't work at all. the ceph command didn't execute properly
[4:49] <dmick> are you running it from a node that has the proper admin keys? Are the keyrings readable as the user you're running it as (do you need tob e root)?
[4:50] * scuttlemonkey (~scuttlemo@c-69-244-181-5.hsd1.mi.comcast.net) has joined #ceph
[4:50] * ChanServ sets mode +o scuttlemonkey
[4:51] <nerdtron> oh wait, i'll try it on the admin node
[4:54] <nerdtron> dmick: no changes..same error
[4:55] <nerdtron> :(
[4:56] <dmick> are your monitors down or something? that's a sick cluster if you can't even ceph to it
[4:56] * scuttlemonkey_ (~scuttlemo@c-69-244-181-5.hsd1.mi.comcast.net) Quit (Ping timeout: 480 seconds)
[4:57] <nerdtron> all mon are up..all osd are up.. i'm really confused on this single pg
[4:57] <dmick> but we're miles from this single pg if you can't even issue a ceph command
[4:57] <dmick> how are you seeing information that this pg is broken?
[4:59] <nerdtron> ceph health detail
[4:59] <nerdtron> everything works in ceph except this pg
[4:59] <nerdtron> w8 luch break.. get back to you. brb
[4:59] <dmick> wait, so ceph health detail currently works
[4:59] <dmick> but the ceph pg query command gives fault?
[5:05] * fireD_ (~fireD@93-142-236-192.adsl.net.t-com.hr) has joined #ceph
[5:06] * iaXe (~axe@223.223.202.194) Quit (Ping timeout: 480 seconds)
[5:07] * fireD (~fireD@93-139-180-225.adsl.net.t-com.hr) Quit (Ping timeout: 480 seconds)
[5:44] <mikedawson> dmick: ping
[5:45] <dmick> pong
[5:47] <mikedawson> I had to rm an osd earlier, now the recovery process has stalled a bit short of completing. I think it is a ceph tunables choose_total_tries issue or something similar. Do you know if there is a good way to kickstart the process again
[5:48] * Vacum_ (~vovo@88.130.217.183) has joined #ceph
[5:48] <dmick> no, not really. recovery problems still by and large stymie me
[5:51] <mikedawson> dmick: ok. thanks. most of the mailing list responses are from gregaf or sjust it seems
[5:55] * Vacum (~vovo@88.130.211.69) Quit (Ping timeout: 480 seconds)
[5:57] <mikedawson> dmick: looks like putting 'tunable choose_local_tries 10' and 'tunable choose_local_fallback_tries 10' in the crushmap (both up from 0) then bouncing the osds got things proceeding again.
[5:57] <dmick> do you have a few PGs not clean, and does query on those PGs hold any clues?
[5:57] <mikedawson> Don't really understand how to track when progress is stopped (or why the code chooses to quit trying to proceed towards HEALTH_OK).
[5:58] <dmick> and...hm
[5:59] <mikedawson> dmick: I feel like having those set to zero should mean "don't quit trying", but it was clearly stuck for some reason
[6:07] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) Quit (Ping timeout: 480 seconds)
[6:08] <skullone> has anyone read up on the seagate object storage drives?
[6:10] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) Quit (Quit: doppelgrau)
[6:12] <mikedawson> skullone: a bit. There has been some thought put in to using them with Ceph. See http://wiki.ceph.com/Planning/Blueprints/Firefly/osd%3A_new_key%2F%2Fvalue_backend
[6:21] * ferai (~quassel@corkblock.jefferai.org) has joined #ceph
[6:25] * miniyo (~miniyo@0001b53b.user.oftc.net) Quit (Ping timeout: 480 seconds)
[6:26] * jefferai (~quassel@corkblock.jefferai.org) Quit (Ping timeout: 480 seconds)
[6:26] * miniyo (~miniyo@0001b53b.user.oftc.net) has joined #ceph
[6:30] <mikedawson> dmick: just found two osds didn't have pids in /var/run/ceph, so they weren't being restarted with the others. Fixed that, and now I'm back in business. 'service ceph restart' didn't show any useful info to alert in the case of a missing pid.
[6:33] * jesus (~jesus@emp048-51.eduroam.uu.se) Quit (Ping timeout: 480 seconds)
[6:37] <dmick> mikedawson: interesting.
[6:38] <nerdtron> all right i think it is not rbd it is mds
[6:38] <nerdtron> mds cluster is degradedmds.ceph-node2 at 10.1.0.12:6800/1060 rank 0 is replaying journal
[6:44] * wschulze (~wschulze@cpe-72-229-37-201.nyc.res.rr.com) Quit (Quit: Leaving.)
[6:47] * jesus (~jesus@emp048-51.eduroam.uu.se) has joined #ceph
[7:10] * Pedras (~Adium@c-67-188-26-20.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[7:17] * nerdtron (~oftc-webi@202.60.8.250) Quit (Quit: Page closed)
[7:34] * nwat (~textual@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Quit: My MacBook has gone to sleep. ZZZzzz???)
[7:37] * mattbenjamin (~matt@76-206-42-105.lightspeed.livnmi.sbcglobal.net) Quit (Quit: Leaving.)
[7:41] * themgt (~themgt@pc-188-95-160-190.cm.vtr.net) Quit (Quit: themgt)
[7:42] * themgt (~themgt@pc-188-95-160-190.cm.vtr.net) has joined #ceph
[7:52] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) Quit (Quit: leaving)
[7:52] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) has joined #ceph
[7:52] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) Quit ()
[8:01] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) has joined #ceph
[8:01] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) Quit ()
[8:01] * capri (~capri@212.218.127.222) has joined #ceph
[8:03] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) has joined #ceph
[8:06] * mikedawson (~chatzilla@c-98-220-189-67.hsd1.in.comcast.net) Quit (Ping timeout: 480 seconds)
[8:10] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) Quit (Quit: leaving)
[8:14] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) has joined #ceph
[8:17] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) Quit ()
[8:19] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) has joined #ceph
[8:20] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) Quit ()
[8:20] * ponyofdeath (~vladi@cpe-75-80-165-117.san.res.rr.com) has joined #ceph
[8:31] * Fetch (fetch@gimel.cepheid.org) Quit (Read error: Connection reset by peer)
[8:36] * Fetch (fetch@gimel.cepheid.org) has joined #ceph
[8:37] * Xedecimal (~Xedecimal@c-24-22-44-186.hsd1.or.comcast.net) Quit (Quit: ~ Trillian - www.trillian.im ~)
[8:48] * Shmouel (~Sam@ns1.anotherservice.com) has joined #ceph
[8:57] * peedu (~peedu@adsl89.uninet.ee) has joined #ceph
[9:03] * rendar (~s@host116-176-dynamic.22-79-r.retail.telecomitalia.it) has joined #ceph
[9:11] * Shmouel1 (~Sam@ns1.anotherservice.com) has joined #ceph
[9:11] * thomnico (~thomnico@2a01:e35:8b41:120:8d9d:271a:75c6:5b0c) has joined #ceph
[9:17] * Shmouel (~Sam@ns1.anotherservice.com) Quit (Ping timeout: 480 seconds)
[9:19] * AfC1 (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) has joined #ceph
[9:19] * yanzheng (~zhyan@jfdmzpr03-ext.jf.intel.com) Quit (Remote host closed the connection)
[9:21] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) Quit (Ping timeout: 480 seconds)
[9:27] * Cube (~Cube@66-87-64-229.pools.spcsdns.net) has joined #ceph
[9:29] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) has joined #ceph
[9:39] * Cube (~Cube@66-87-64-229.pools.spcsdns.net) Quit (Quit: Leaving.)
[9:41] * jcsp (~Adium@0001bf3a.user.oftc.net) Quit (Read error: Operation timed out)
[9:50] * garphy`aw is now known as garphy
[10:03] * julian (~julianwa@125.70.133.91) has joined #ceph
[10:18] * ScOut3R (~scout3r@540205B4.dsl.pool.telekom.hu) has joined #ceph
[10:21] * LeaChim (~LeaChim@host86-161-89-52.range86-161.btcentralplus.com) has joined #ceph
[10:27] * HVT (~root@118.70.170.126) has joined #ceph
[10:28] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) has joined #ceph
[10:29] <HVT> hi guys
[10:30] <HVT> im developing some apps using librados from ceph
[10:30] * AfC1 (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) Quit (Ping timeout: 480 seconds)
[10:31] <HVT> im wondering how could i compile the librados only instead of the whole enormous ceph package?
[10:37] * i_m (~ivan.miro@95.180.8.206) has joined #ceph
[10:41] * thomnico (~thomnico@2a01:e35:8b41:120:8d9d:271a:75c6:5b0c) Quit (Quit: Ex-Chat)
[10:48] * mozg (~andrei@46.229.149.194) has joined #ceph
[10:53] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) has joined #ceph
[10:54] * HVT1 (~root@14.160.47.146) has joined #ceph
[10:55] * HVT (~root@118.70.170.126) Quit (Read error: Connection reset by peer)
[11:10] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) Quit (Quit: tryggvil)
[11:10] * djezz (~jasper.si@target15.rcitlab.rug.nl) Quit (Read error: Connection reset by peer)
[11:10] * timmeah (~tim@target15.rcitlab.rug.nl) Quit (Read error: Connection reset by peer)
[11:18] * zidarsk8 (~zidar@prevod.fri1.uni-lj.si) has joined #ceph
[11:19] * zidarsk8 (~zidar@prevod.fri1.uni-lj.si) has left #ceph
[11:24] * Discard (~discard@APuteaux-651-1-98-51.w82-124.abo.wanadoo.fr) has joined #ceph
[11:24] <Discard> hi there
[11:24] <Discard> I have a problem with RGW
[11:26] <Discard> when I do a wget http://p1.13h.com/swift/v1/test/test.mp3 there is no problem to get the mp3 etcetc ...
[11:26] <Discard> but when i put this url in a browser or in VLC
[11:27] <Discard> it stopped playing this file
[11:27] <Discard> anyone got an idea ?
[11:35] <ccourtaut> Discard: works on my machine with chromium and VLC
[11:38] <Discard> ccourtaut: bad news for me ! could you listen to full track ?
[11:38] <Discard> because it stoppes after 5 to 32 seconds
[11:39] <ccourtaut> Discard: i seeked into the track and seemed to work
[11:39] <ccourtaut> can't listen to the whole track, in openspace, bluetooth broken in my distrib so headset not working
[11:40] <ccourtaut> Discard: which version of ceph are you running?
[11:40] <Discard> 72.2
[11:40] <Discard> 0.72.2
[11:41] <ccourtaut> ok
[11:43] <Discard> safari detects a live streaming not a full mp3.
[11:44] * mschiff (~mschiff@port-19476.pppoe.wtnet.de) has joined #ceph
[11:45] <Discard> It is as if the gateway doesn't send file headers
[11:48] <Vacum_> Discard: but it does: http://pastie.org/8596539
[11:48] * Vacum_ is now known as Vacum
[11:50] <Discard> Vacum: ok a wget is ok but if you are on mac os X with chrome it stoppes after 5 secs
[11:51] <Vacum> Discard: I just wanted to point out that the Content-Type is set correctly
[11:51] <Vacum> Discard: if you download the file and then throw it into chrome locally, does it play completely? perhaps the file is broken?
[11:52] * Shmouel1 (~Sam@ns1.anotherservice.com) Quit (Ping timeout: 480 seconds)
[11:52] <Discard> Vacum: if I put the same file on local or on an apache session it's ok too
[11:53] <Discard> on windows it's the same !
[11:55] <Vacum> Discard: when I download that file from your server, it looks like this: http://pastie.org/8596562
[11:56] <Vacum> Discard: whatever you did, the file contains some old HTTP header garbage at its beginning
[11:56] <Vacum> from 2th of Jan 8:31:15 GMT
[11:56] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[11:56] <Vacum> Discard: which is yesterday, so its not from my wget call :)
[11:57] * ScOut3R (~scout3r@540205B4.dsl.pool.telekom.hu) Quit (Remote host closed the connection)
[11:57] * ScOut3R (~scout3r@540205B4.dsl.pool.telekom.hu) has joined #ceph
[11:58] <Discard> Vacum: ouch :-) thank you very much :-)
[11:58] <Vacum> welcome :)
[11:58] <Discard> I think there a problem when i upload it on the gatewy
[12:02] * allsystemsarego (~allsystem@5-12-241-225.residential.rdsnet.ro) has joined #ceph
[12:03] * Shmouel (~Sam@fny94-12-83-157-27-95.fbx.proxad.net) has joined #ceph
[12:03] * Shmouel (~Sam@fny94-12-83-157-27-95.fbx.proxad.net) Quit ()
[12:04] * Shmouel (~Sam@ns1.anotherservice.com) has joined #ceph
[12:08] * Discard_ (~discard@213-245-29-151.rev.numericable.fr) has joined #ceph
[12:09] * HVT1 (~root@14.160.47.146) has left #ceph
[12:12] * diegows (~diegows@190.190.17.57) has joined #ceph
[12:12] * shang (~ShangWu@175.41.48.77) Quit (Quit: Ex-Chat)
[12:12] * shang (~ShangWu@175.41.48.77) has joined #ceph
[12:13] * Discard (~discard@APuteaux-651-1-98-51.w82-124.abo.wanadoo.fr) Quit (Ping timeout: 480 seconds)
[12:13] * Discard_ is now known as Discard
[12:13] * shang (~ShangWu@175.41.48.77) Quit ()
[12:13] * shang (~ShangWu@175.41.48.77) has joined #ceph
[12:15] * yanzheng (~zhyan@134.134.137.75) has joined #ceph
[12:22] <Discard> Vacum: it's very strange when i do a swift command on objects, I have no problem ... it's only when a use HTTP directly
[12:26] <Discard> on this wget http://p1.13h.com/swift/v1/test/touslesmemes.mp3
[12:26] <Discard> there is no strange headers
[12:30] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[12:33] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[12:39] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[12:47] * shang (~ShangWu@175.41.48.77) Quit (Ping timeout: 480 seconds)
[12:47] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[12:47] * andreask (~andreask@h081217067008.dyn.cm.kabsi.at) has joined #ceph
[12:47] * ChanServ sets mode +v andreask
[13:03] * yeled (~yeled@spodder.com) Quit (Quit: meh..)
[13:04] * yeled (~yeled@spodder.com) has joined #ceph
[13:04] * yeled (~yeled@spodder.com) Quit ()
[13:05] * yeled (~yeled@spodder.com) has joined #ceph
[13:10] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[13:13] * AfC (~andrew@2001:44b8:31cb:d400:6e88:14ff:fe33:2a9c) Quit (Quit: Leaving.)
[13:15] * yeled (~yeled@spodder.com) Quit (Quit: .)
[13:16] * andreask (~andreask@h081217067008.dyn.cm.kabsi.at) Quit (Quit: Leaving.)
[13:17] * andreask (~andreask@h081217067008.dyn.cm.kabsi.at) has joined #ceph
[13:17] * ChanServ sets mode +v andreask
[13:19] * yeled (~yeled@spodder.com) has joined #ceph
[13:28] * garphy is now known as garphy`aw
[13:34] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[13:35] * rudolfsteiner (~federicon@181.167.96.123) Quit ()
[13:36] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[13:37] <Meyer^> -*????????.,,,,,,,,,,,,
[13:37] <Discard> ?
[13:37] <Meyer^> '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
[13:37] <Meyer^> gah
[13:37] <Meyer^> Discard: cat on keyboard
[13:40] <janos> haha
[13:41] * niklas (niklas@vm15.hadiko.de) Quit (Ping timeout: 480 seconds)
[13:44] * Midnightmyth (~quassel@93-167-84-102-static.dk.customer.tdc.net) has joined #ceph
[13:44] * ksingh (~Adium@2001:708:10:10:b0f7:328e:c1b5:a5f1) has joined #ceph
[13:44] * zidarsk8 (~zidar@84-255-203-33.static.t-2.net) has joined #ceph
[13:45] * zidarsk8 (~zidar@84-255-203-33.static.t-2.net) has left #ceph
[13:45] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[13:46] <ksingh> cephers need advice on cephfs ==> i have cephfs working good but as soon as i mount ceph filesystem it shows my entire ceph cluster size
[13:47] <ksingh> 192.168.200.101:6789:/ 419T 8.5T 411T 3% /cephfs
[13:48] <ksingh> what i want is ==> i have a pool created for cephfs with name "cephfs_pool1" , i want to mount this pool
[13:48] <ksingh> how to mount this pool ???
[13:49] * jcsp (~jcsp@0001bf3a.user.oftc.net) has joined #ceph
[13:50] * niklas (niklas@home.hadiko.de) has joined #ceph
[13:54] <Discard> any tuto to use Lighttpd instead of apache ?
[13:55] <Discard> on Object Gateway ?
[13:58] <mozg> Discard, yeah, I would also like to know that. Apache is too heavy
[13:58] <mozg> I've tried to use nginx instead of apache and it doesn't properly work
[13:59] <mozg> it's got an issue with large file size copy, which apache doesn't have
[13:59] <janos> i used to use lighttpd for other things - is that still maintained?
[13:59] <mozg> but for small files, nginx uses about 1/6th of resources compared to apache
[14:00] <mozg> Has anyone successfully configured nginx as the proxy for radosgw?
[14:01] * jcsp (~jcsp@0001bf3a.user.oftc.net) Quit (Ping timeout: 480 seconds)
[14:01] * alphe (~alphe@0001ac6f.user.oftc.net) has joined #ceph
[14:02] * rudolfsteiner (~federicon@181.167.96.123) has joined #ceph
[14:10] * rudolfsteiner (~federicon@181.167.96.123) Quit (Quit: rudolfsteiner)
[14:16] * hjjg (~hg@p3EE309A3.dip0.t-ipconnect.de) has joined #ceph
[14:23] * Sysadmin88 (~IceChat77@2.218.8.40) has joined #ceph
[14:24] * Sysadmin88 (~IceChat77@2.218.8.40) Quit ()
[14:24] * Sysadmin88 (~IceChat77@2.218.8.40) has joined #ceph
[14:27] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) has joined #ceph
[14:29] * Sysadmin88 (~IceChat77@2.218.8.40) Quit (Quit: On the other hand, you have different fingers.)
[14:31] * t0rn (~ssullivan@2607:fad0:32:a02:d227:88ff:fe02:9896) has joined #ceph
[14:32] * jesus (~jesus@emp048-51.eduroam.uu.se) Quit (Ping timeout: 480 seconds)
[14:37] * peedu (~peedu@adsl89.uninet.ee) Quit (Read error: Connection reset by peer)
[14:41] * KevinPerks (~Adium@cpe-066-026-252-218.triad.res.rr.com) has joined #ceph
[14:43] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) Quit (Ping timeout: 480 seconds)
[14:44] * jesus (~jesus@emp048-51.eduroam.uu.se) has joined #ceph
[14:45] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) has joined #ceph
[14:46] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) Quit ()
[14:46] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) has joined #ceph
[14:46] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) Quit ()
[14:46] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) has joined #ceph
[14:47] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) Quit ()
[14:47] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) has joined #ceph
[14:48] * jcsp (~jcsp@0001bf3a.user.oftc.net) has joined #ceph
[14:50] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) Quit ()
[14:51] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) has joined #ceph
[14:53] * BillK (~BillK-OFT@106-69-25-13.dyn.iinet.net.au) Quit (Ping timeout: 480 seconds)
[14:58] * shang (~ShangWu@220-135-203-169.HINET-IP.hinet.net) Quit (Quit: Ex-Chat)
[15:00] * thomnico (~thomnico@AMarseille-652-1-314-250.w90-8.abo.wanadoo.fr) has joined #ceph
[15:00] * mozg (~andrei@46.229.149.194) Quit (Remote host closed the connection)
[15:01] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) has joined #ceph
[15:06] <Discard> anyone has an idea for this bug ?
[15:06] <Discard> http://pastebin.com/evRsaBxm
[15:09] * jcsp (~jcsp@0001bf3a.user.oftc.net) Quit (Quit: Leaving.)
[15:09] * jcsp (~Adium@0001bf3a.user.oftc.net) has joined #ceph
[15:11] * CooLCaT (CooLCaT@chello213047004235.west1.11.vie.surfer.at) has joined #ceph
[15:12] <CooLCaT> good afternoon
[15:12] <alphe> CooLCaT hello
[15:12] <CooLCaT> may i ask something about replication level and failure of one storage node?
[15:13] <CooLCaT> we have 3 storage nodes each with 4 osd
[15:13] <CooLCaT> replication level is set to 3
[15:14] <CooLCaT> so the object lies on one osd on each node
[15:14] <CooLCaT> what happens if one host goes down?
[15:15] <janos> then you essentially have 2 copies instead of 3
[15:16] <CooLCaT> ok
[15:17] <CooLCaT> so now i have to think to translate from german to english what i want to know :)
[15:17] <janos> haha
[15:17] <janos> i won't necessarily know, but i will answer to the best of my knowledge
[15:17] <CooLCaT> as replicalevel is 3 will the cluster start to make a third copy on the remaining two nodes
[15:18] <janos> with a failure domain of "host" i would say no
[15:18] <janos> but i do not know 100%
[15:18] <janos> it will remain in a degraded state
[15:18] <CooLCaT> because we dont know what we should set as near full and full ratio
[15:20] <alphe> coolcat near full and full ratio are indice to warn you and start discarding nodes before they get 100 full
[15:20] <alphe> is a disk is 100% full in linux it can t be operated anymore
[15:21] <alphe> that is why you have to set warn for near full and full ratio below 100%
[15:21] <CooLCaT> ok i think i get the picture
[15:21] <alphe> the near full warn allows you to change the weight of the osd in the osd in order to not make it as priority to be write on than others
[15:22] * andreask (~andreask@h081217067008.dyn.cm.kabsi.at) has left #ceph
[15:22] <CooLCaT> if we set replica leve 3,failure domain host and min size to 2,then if two nodes fail completly the cluster stops reading/writing but we still have and copy on the remaining node
[15:22] <alphe> it is indicative that you need to extend your ceph cluster adding new nodes
[15:22] <alphe> or remove files
[15:23] <CooLCaT> :) i guess the missing link was the failure domain :)
[15:23] <CooLCaT> we always thought,that if one node goes completly down, the replica level would be kept up in the remaining two nodes
[15:23] <alphe> coolcat yes that will allow your nodes to be restored as they where when you enter 2 new nodes
[15:24] <alphe> but a 66% node faillure is heavy ... and should never happend reconstructing 66% of your cluster capacity will have a heavy impact on your cluster for a long period of time
[15:25] <alphe> not telling they you will loose 66% of your global space to replicas
[15:25] <CooLCaT> we surely hope that that would never happen.was some kind of worst case
[15:25] <alphe> if you have 10TB that means 6TB are replicas and only 3TB is usable ...
[15:26] <alphe> even a bit less than ?3TB
[15:26] <alphe> CooLCaT best way to not be so "faillure" dependent in fact is multiplying the physical nodes of your cluster
[15:27] <alphe> make a hardware faillure 1 of a thousand and you will have a rock stable and fast pace recovering ceph cluster
[15:28] <CooLCaT> mhm your right.that is one thing we discussed that replevel 3 means roughly 1/3 of total capacity available
[15:28] <alphe> if you have a lot of physical nodes then you can set your replicas to a lower number like 2 and enjoy almost half of your overall data store capacity
[15:28] * mkoo (~mkoo@chello213047069005.1.13.vie.surfer.at) has joined #ceph
[15:28] <CooLCaT> hi mike
[15:29] <alphe> which is better than enjoying a less than a third of it
[15:29] <CooLCaT> we are starting with three nodes for now
[15:29] <alphe> CooLCaT you can do really really cheap nodes...
[15:30] <alphe> actually the nodes I use are less than 200 USD ... and working quite well if I don t mess up the inicial installation
[15:30] <CooLCaT> :)
[15:30] <janos> alphe: really? you have links?
[15:30] <janos> alphe: i'd be interested in that for home
[15:30] <alphe> janos I build them :P want to buy some ?
[15:30] <mkoo> 200bucks?! thats a treat
[15:30] <janos> alphe: depending on what they are, sure!
[15:31] <janos> you have piqued my curiosity
[15:31] <alphe> janos basically there is no magic for that price you can get atom D2500 CCE motherboard 2Gb of ram and 2 2TB drives
[15:32] <alphe> more or less the thing that cost us the most is the box that hold those blades
[15:32] <CooLCaT> hmm d2500 cee 80 Euros
[15:32] <janos> yeah i was wondering about case
[15:32] <janos> the enclosures get more expensive at that point
[15:33] <alphe> janos not that much it arround 2000 USD for 10 blades that makes the whole thing around 4000 which is the price of en entry server blade of HP proliant line with 2 small disks
[15:33] <alphe> or even no disks at all
[15:34] <janos> not bad if it has the horsepower to be and osd
[15:34] <alphe> CooLCaT that is why it is better to buy them per 50 motherboard than by one ...
[15:35] <CooLCaT> http://www.buildablade.com/bb-itx96.htm
[15:35] <alphe> janos well it is fairly descent I never stuff the nodes or get them extra laggy
[15:35] <alphe> and I get on single files data transfer rate over 100 MB in LAN
[15:35] <janos> CooLCaT: haha cool
[15:36] <CooLCaT> :) just dont know where to get it in europe
[15:36] <doppelgrau> janos: I've bought me for testing purpose four zotac id18 - 110 EUR, but only one internal 2,5' hdd. With 4GB of RAM and a small disk less than 200 EUR
[15:36] <alphe> it is something like that I do
[15:36] <CooLCaT> ok found something 1199$ - phew
[15:37] <janos> doppelgrau: yeah those zotac units are nice. i have one of their models hooked to my tv with openelec on it
[15:37] <janos> it's maybe 4 inches square
[15:38] <alphe> CooLCaT actually we are still in developpement state but we like the 10 nodes way
[15:41] * markbby (~Adium@168.94.245.4) has joined #ceph
[15:43] <CooLCaT> altough that would be 133 per blade
[15:46] <alphe> actually with the actuall design my worst problem is the os being on usb pendrives ...
[15:47] <alphe> if a monitor service run on one of those pendrive that strain it too much and it often clash ...
[15:48] <alphe> clash means the kernel detects IO errors and remount the filesystem partition in readonly mode which as no impact on the osd ... but that ends crashing the mon service that has no more rights to write files in it s "store database"
[15:49] <alphe> so preferencially having a 11th machine to host the main monitor is a must have
[15:49] <janos> yeah i would design mons with more durability than osd's
[15:49] <alphe> but that machine too can be a atom based machine ... only trick is that it need to have a real hard drive
[15:50] <alphe> janos the log part of the mon service in ceph from experiments reveal to be really agressive and strainfull
[15:51] <alphe> janos it can represent a dayly size of 1GB... if your filesystem os particion is about 5GB that is really a probleme ...
[15:51] * portante is now known as portante|afk
[15:51] <alphe> but a running working ceph node on a 2 to 4 euros pendrive it is really a cool option
[15:51] <janos> yeah
[15:52] <alphe> it the node crash you can easyly run fsck on it and dd the content to a new pendrive
[15:53] <alphe> or better have a set of spare pendrives already installed with os and ceph same version and simply deploy /var/lib/ceph/ files to that new pendrive and in a couple of minutes you have your node back to life
[15:54] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) has left #ceph
[15:55] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) has joined #ceph
[15:55] <alphe> janos one option I will see now is that an LVM or anyother kind of software raid can help me solve the pendrive faillures reducing the strain on the pendrive
[15:56] <alphe> so instead of 1 8gb pendrive I would put 2 and if one fail then simply replace the failled
[15:56] <janos> there are some rather fast pendrives out these days
[15:56] <janos> they cost more though
[15:56] <alphe> janos yes
[15:57] <alphe> janos not that much
[15:57] <janos> i haven't looked into how many expected writes they have
[15:57] <alphe> the pendrive i use are USB 2.0 and stamped HI speed
[15:57] <janos> on a Mon i would probably still do a drive. possibly mSATA
[15:57] <alphe> they work fairely right ...
[15:59] <alphe> well with all that info I will encourage competition and people to do same stuff as I do but I don t fear that ... I'm coonvinced that the most of the work is in service more than on machine ...
[15:59] <alphe> a great support is way better than a great machine in my opinion ...
[15:59] <janos> great support is becoming rare
[16:00] <alphe> easy to administrate ceph easy to maintain hardware is the key designs
[16:00] <janos> so if you can do it, i htink that's a good thing
[16:00] <alphe> janos yes ...
[16:01] <alphe> janos it is more easy to take client for the fools they are cheat them and tigh them with stupid technologies than doing the stuff the right way
[16:01] <janos> sadly
[16:02] <alphe> but I think ceph is going the right way I want to be part of that concient good way to do business even if it means investing a lot on staff formation building trust with clients and being a bit expensive
[16:03] <alphe> a bit expensive not that much if you compare with oracle or netapp ...
[16:04] <janos> not much looks expensive net to them haha
[16:04] <janos> net/next
[16:04] <alphe> or even a bunch of dell R520 poweredge servers
[16:05] <alphe> if you buy 5 poweredge servers you have too much horse power too few disks hosted and you are investing what 3 or 4 of my boxes cost ...
[16:05] <alphe> 1 of my box = 40TB of global storare 18TB of usable confortable storage
[16:07] <janos> yeah
[16:07] <janos> it sounds like a decent plan
[16:08] <CooLCaT> just doing a bit on searching for itx blade chassis. one problem imho is that you can install only 2 drives per blade
[16:09] <alphe> CooLCaT yes that is our disign
[16:09] <doppelgrau> CooLCaT: buildablade offers 2*2 drives per Blade if I see ist allright
[16:09] <alphe> design that is way we focus the hardrives for osd purpose and the os in in a pendrive
[16:10] <Vacum> alphe: so you put 2 * 2TB SATA drives to one Atom CPU and 4GB RAM?
[16:10] <alphe> that way you have 4TB in 2 disks of 2TB per node and 2 osd per node
[16:10] <alphe> vacum 2 gb of ram is enough ... and trust me my ceph cluster eat 2TB a day
[16:10] <Vacum> alphe: and 10 of those nodes into one blade chassis of how many unit heights?
[16:10] <alphe> 8U
[16:11] <alphe> but spendidly ventilated
[16:11] <Vacum> and those nodes have GBit I assume?
[16:11] <CooLCaT> i once tried to install openfiler on a pen drive - didnt work as expected,so i am not a friend of putting the system of a pen drive
[16:11] <CooLCaT> of/on
[16:12] <alphe> at full power the cpu are at 40 degrees constantly in a room of 30 degree (yes my boss hate investing in air cooling system he preferes buy ceph nodes"
[16:12] <CooLCaT> lol
[16:12] <alphe> Vacum yes 2 gbit ethernet
[16:12] <Vacum> alphe: so you are rougly at 5TB gross storage per Unit Height?
[16:12] <kraken> http://i.imgur.com/XEEI0Rn.gif
[16:13] <alphe> but this shows that the box 8U we have is pretty ok for cooling down 10 blades of 20 desktop regular hard drives
[16:13] <alphe> vacum 4TB
[16:13] <alphe> a bit less because of xfs and rbd
[16:13] <Vacum> alphe: 40TB/8U = 5TB/U ?
[16:14] <alphe> yes if you count per U
[16:14] <alphe> but if that is your concern I saw a company proposing a ceph store with 60 disk in 6U
[16:14] <Vacum> and one 8U box fully equipped ends up at $... ? roughly?
[16:14] <alphe> that made me wanting so much to buy one ...
[16:15] * pmatulis (~peter@64.34.151.178) has joined #ceph
[16:16] <alphe> Vacum with disk motherboard pendrives box switch cables etc around 5000 USB
[16:17] <alphe> but it is hard to tell because the most you buy at once the more you can lower prices ...
[16:17] <alphe> and what cost us the most in all this is the transport since we are not crafting the box
[16:17] <Vacum> not bad. that is $125/TB gross
[16:17] <kraken> http://i.imgur.com/XEEI0Rn.gif
[16:18] <Vacum> mh. is kraken a bot that triggers on the word gross ?
[16:18] <kraken> http://i.imgur.com/XEEI0Rn.gif
[16:18] <Vacum> it is
[16:18] <alphe> we could try to get local crafter involved that would half the price more or less for our local client
[16:20] <alphe> but for the moment we really seek to fullfill our internal need with the prototype going through all kind of problems to build experience and on top of that I still hope to get some free time soon to get started the project of writing a distributed client for windows that directly speaks to rbd
[16:20] <alphe> ceph
[16:20] <Vacum> 2GB for two OSDs can become an issue if a lot of small files are stored on the OSDs, as XFS can take itself a lot of kernel slabs for inodes. I had a box using ~28GB kernel slabs only for xfs innodb
[16:21] <Vacum> s/innodb/inodes
[16:21] <kraken> Vacum meant to say: 2GB for two OSDs can become an issue if a lot of small files are stored on the OSDs, as XFS can take itself a lot of kernel slabs for inodes. I had a box using ~28GB kernel slabs only for xfs inodes
[16:22] <Vacum> nice feature
[16:22] <Vacum> !help kraken
[16:22] <kraken> Sorry Vacum, I don't know about that plugin
[16:22] <Vacum> whos bot is that?
[16:23] * wschulze (~wschulze@cpe-72-229-37-201.nyc.res.rr.com) has joined #ceph
[16:24] <alphe> vacum until now i did see any problems of memory saturation
[16:24] <Vacum> alphe: ok, great
[16:24] <alphe> I noticed a tendency of using swap but that is all and it is around 200MB for a 2GB swap
[16:25] <alphe> but the fact is ceph fastastically distribute needs over ressources so the more ressources you have the best it works
[16:26] <alphe> my plan is to reahc production level rapidely and then dismantle the RAID SAS enclosure based monolitic store we already have and recycle them into new ceph boxes as soon as posible
[16:26] <Vacum> alphe: any reason why you only use 2TB drives and not 3TB? 3TB has currently the sweet spot for $/TB
[16:26] <alphe> extending the ceph store capabilities
[16:28] <alphe> VAcum atom D2500CCE motherboard don t allow SATA 3 so 4tb with sata 2 is slower not sure that the bios with manage 4tb drives anyway and 2tb drive where already bought for a previous ceph prototype that crashed and burnt
[16:28] <alphe> crash and burnt litterally it was a poor design and a cheap solution never worked
[16:28] <Vacum> wow
[16:28] <alphe> because yes a cheaper solution is still possible than this one
[16:29] <Vacum> sure, you could use cardboard boxes instead of metal chassis...
[16:30] <alphe> vacum yes
[16:30] <alphe> Vacum or use raspeberry like :)
[16:30] <Vacum> :)
[16:31] <alphe> vacum calxeda company proposed a 60 disks in 4 U storage
[16:31] <alphe> I try to make contact with them never gets a single reply
[16:32] <Vacum> ah, the ARM based thing?
[16:32] * dmsimard (~Adium@108.163.152.2) has joined #ceph
[16:33] <alphe> calxeda cahnged their website they removing all the fancy great looking an appealing pictures from it ...
[16:33] * mnash (~chatzilla@66-194-114-178.static.twtelecom.net) Quit (Read error: Connection reset by peer)
[16:34] * mnash (~chatzilla@vpn.expressionanalysis.com) has joined #ceph
[16:34] <Discard> hi there anyone use lighted instead of apache ?
[16:34] <alphe> Vacum yes
[16:34] <alphe> Discard nope
[16:35] <alphe> discard for s3-amazon like distributed storage over ceph ?
[16:35] <Discard> alphe: swift yep :-) i have a problem with apache
[16:36] <Discard> http://pastebin.com/evRsaBxm
[16:36] <Discard> so I tried other http server
[16:36] <alphe> discard the official documentation is all about apache (and have lacks and missing things like activating the SSL layer properly)
[16:37] <alphe> dsicard utiliser ceph pour stocker des mp3 c est mal hadopi va te faire la tronche :P
[16:37] <alphe> hehehe
[16:38] <Discard> alphe: c'est un projet startup :-)
[16:38] <alphe> discard tu fais comme tu veux :)
[16:38] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) Quit (Remote host closed the connection)
[16:38] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) has joined #ceph
[16:38] <Discard> ce qu eue veux dire c'est que nous avons les droits sacem
[16:39] <Discard> bref
[16:39] <alphe> hum that is weird the session i properly negociated but nothing is sent
[16:39] * dalchand (~dalchand@122.172.6.11) has joined #ceph
[16:39] * ksingh (~Adium@2001:708:10:10:b0f7:328e:c1b5:a5f1) Quit (Quit: Leaving.)
[16:40] <Discard> yep it's only with a browser web ! if I do a wget there is no error ...
[16:40] <alphe> hum seems like a problem of right
[16:40] <Discard> I think I have to tune something but I don't now what :P
[16:41] <Discard> which right ?
[16:41] <alphe> the read rights ...
[16:41] <alphe> on the ceph cluster
[16:42] <alphe> but that is probably not the reason ...
[16:42] <Discard> how is possible I can download file with a wget command
[16:42] <Discard> on the same server
[16:42] * mattbenjamin (~matt@aa2.linuxbox.com) has joined #ceph
[16:42] <alphe> dsicard good question ... I am no swift specialist .
[16:43] <alphe> dsicard can you try another web browser ?
[16:43] <Discard> yep
[16:43] <alphe> and it works ?
[16:43] <Discard> nope
[16:43] <kraken> http://i.imgur.com/xKYs9.gif
[16:43] <Discard> with safari it's ok
[16:43] <dalchand> is there any tool to browse object store... ?
[16:44] <dalchand> im using swift interface
[16:44] <Discard> dalchand: cyberduck on mac
[16:44] <dalchand> will it work on ubuntu?
[16:44] <alphe> dalchand -> google swift client for my os
[16:44] <alphe> dalchand nope
[16:44] <dalchand> okk
[16:45] * nwat (~textual@c-50-131-197-174.hsd1.ca.comcast.net) has joined #ceph
[16:46] <dalchand> is there any one who is using s3 interface
[16:46] <dalchand> i tried ... it didn't work
[16:46] <alphe> cyberduck has a windows version
[16:46] <dalchand> i'm able to authenticate but upon creating bucket it says 405 MethodNotAllowed
[16:47] <alphe> dalchand I used a perl script tool to access my s3 amazon storage create the bucket and send retrive files
[16:47] <alphe> but s3 appeared to stop working after an update
[16:48] <dalchand> have you faced that problem?
[16:48] * diegows (~diegows@190.190.17.57) Quit (Ping timeout: 480 seconds)
[16:48] <alphe> nope I went cephfs and ended getting more problems so now I go rbd
[16:48] <alphe> and I hope that will work properly
[16:49] * yanzheng (~zhyan@134.134.137.75) Quit (Remote host closed the connection)
[16:50] <dalchand> they say it could be because of missing dns configuration for amazon like subdomains but i have configured it and im able to ping any subdomain *.aws.mydomain.com
[16:56] <alphe> have to leave have a good day all
[16:56] * alphe (~alphe@0001ac6f.user.oftc.net) Quit (Quit: Leaving)
[16:57] * diegows (~diegows@200.68.116.185) has joined #ceph
[16:58] <pmatulis> hago
[16:58] * xcrracer (~xcrracer@fw-ext-v-1.kvcc.edu) has joined #ceph
[17:03] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) has joined #ceph
[17:06] * jcsp1 (~Adium@82-71-55-202.dsl.in-addr.zen.co.uk) has joined #ceph
[17:07] * perfectsine (~perfectsi@if01-gn01.dal05.softlayer.com) Quit (Quit: Leaving)
[17:07] * perfectsine (~perfectsi@if01-gn01.dal05.softlayer.com) has joined #ceph
[17:12] * jcsp (~Adium@0001bf3a.user.oftc.net) Quit (Ping timeout: 480 seconds)
[17:16] * elder (~elder@c-71-195-31-37.hsd1.mn.comcast.net) Quit (Quit: Leaving)
[17:16] * Steki (~steki@fo-d-130.180.254.37.targo.rs) has joined #ceph
[17:19] * Cube (~Cube@66-87-64-229.pools.spcsdns.net) has joined #ceph
[17:21] * bandrus1 (~Adium@107.222.158.227) Quit (Ping timeout: 480 seconds)
[17:22] * BManojlovic (~steki@fo-d-130.180.254.37.targo.rs) Quit (Ping timeout: 480 seconds)
[17:26] * elder (~elder@c-71-195-31-37.hsd1.mn.comcast.net) has joined #ceph
[17:26] * ChanServ sets mode +o elder
[17:38] * julian_ (~julianwa@125.70.132.28) has joined #ceph
[17:39] * bandrus (~Adium@c-98-238-176-251.hsd1.ca.comcast.net) has joined #ceph
[17:41] * hjjg_ (~hg@p3EE31DE7.dip0.t-ipconnect.de) has joined #ceph
[17:41] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) has left #ceph
[17:41] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) has joined #ceph
[17:43] * hjjg (~hg@p3EE309A3.dip0.t-ipconnect.de) Quit (Ping timeout: 480 seconds)
[17:45] * julian (~julianwa@125.70.133.91) Quit (Ping timeout: 480 seconds)
[17:49] * julian_ (~julianwa@125.70.132.28) Quit (Quit: afk)
[17:51] * portante|afk is now known as portante
[17:53] * Pedras (~Adium@c-67-188-26-20.hsd1.ca.comcast.net) has joined #ceph
[17:54] * ksingh (~Adium@91-145-88-160.bb.dnainternet.fi) has joined #ceph
[17:56] * bjornar (~bjornar@ti0099a340-dhcp0395.bb.online.no) has joined #ceph
[17:58] * Shmouel (~Sam@ns1.anotherservice.com) Quit (Ping timeout: 480 seconds)
[17:59] * dalchand (~dalchand@122.172.6.11) Quit (Quit: Leaving)
[18:02] * thomnico (~thomnico@AMarseille-652-1-314-250.w90-8.abo.wanadoo.fr) Quit (Quit: Ex-Chat)
[18:03] * ksingh (~Adium@91-145-88-160.bb.dnainternet.fi) Quit (Quit: Leaving.)
[18:16] * mikedawson (~chatzilla@2602:306:c550:6c49:c83b:b190:5dd0:dec8) has joined #ceph
[18:18] * mikedawson (~chatzilla@2602:306:c550:6c49:c83b:b190:5dd0:dec8) Quit ()
[18:23] * bjornar (~bjornar@ti0099a340-dhcp0395.bb.online.no) Quit (Ping timeout: 480 seconds)
[18:26] * rendar (~s@host116-176-dynamic.22-79-r.retail.telecomitalia.it) Quit ()
[18:32] * rudolfsteiner (~federicon@200.68.116.185) has joined #ceph
[18:36] <Discard> is there an option to put radiogw daemon more verbose ?
[18:37] * tryggvil (~tryggvil@17-80-126-149.ftth.simafelagid.is) Quit (Quit: tryggvil)
[18:38] * thomnico (~thomnico@AMarseille-652-1-314-250.w90-8.abo.wanadoo.fr) has joined #ceph
[18:38] * ircolle (~Adium@2601:1:8380:2d9:780e:1e0c:6b33:581d) has joined #ceph
[18:40] * xarses (~andreww@c-24-23-183-44.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[18:42] * rudolfsteiner (~federicon@200.68.116.185) Quit (Quit: rudolfsteiner)
[18:42] * Underbyte (~jerrad@pat-global.macpractice.net) Quit (Quit: Linkinus - http://linkinus.com)
[18:45] * sagelap (~sage@125.sub-70-197-68.myvzw.com) has joined #ceph
[18:45] <pmatulis> Discard: there should be a logging subsystem for it that you can dial up
[18:47] <Discard> :-)
[18:47] * angdraug (~angdraug@12.164.168.115) has joined #ceph
[18:48] * joshd1 (~jdurgin@2602:306:c5db:310:60c7:5c33:4fc1:8547) has joined #ceph
[18:51] * thomnico (~thomnico@AMarseille-652-1-314-250.w90-8.abo.wanadoo.fr) Quit (Ping timeout: 480 seconds)
[18:52] * hjjg_ (~hg@p3EE31DE7.dip0.t-ipconnect.de) Quit (Ping timeout: 480 seconds)
[19:05] * mschiff_ (~mschiff@port-91637.pppoe.wtnet.de) has joined #ceph
[19:12] * mschiff (~mschiff@port-19476.pppoe.wtnet.de) Quit (Ping timeout: 480 seconds)
[19:14] * xarses (~andreww@12.164.168.115) has joined #ceph
[19:18] * Tamil (~tamil@38.122.20.226) has left #ceph
[19:19] * Tamil (~tamil@38.122.20.226) has joined #ceph
[19:28] * Pedras (~Adium@c-67-188-26-20.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[19:30] * bandrus1 (~Adium@108.246.12.107) has joined #ceph
[19:33] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) has joined #ceph
[19:33] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) Quit (Quit: Computer has gone to sleep.)
[19:35] * bjornar (~bjornar@ti0099a340-dhcp0395.bb.online.no) has joined #ceph
[19:35] * Cube (~Cube@66-87-64-229.pools.spcsdns.net) Quit (Quit: Leaving.)
[19:36] * bandrus (~Adium@c-98-238-176-251.hsd1.ca.comcast.net) Quit (Read error: Operation timed out)
[19:42] * i_m (~ivan.miro@95.180.8.206) Quit (Quit: Leaving.)
[19:42] * i_m (~ivan.miro@95.180.8.206) has joined #ceph
[19:51] * ksingh (~Adium@85-76-182-225-nat.elisa-mobile.fi) has joined #ceph
[19:56] * comptona (~oftc-webi@72.5.114.17) has joined #ceph
[19:56] * comptona (~oftc-webi@72.5.114.17) Quit ()
[19:59] * sagelap1 (~sage@38.122.20.226) has joined #ceph
[20:04] * psomas (~psomas@inferno.cc.ece.ntua.gr) has joined #ceph
[20:06] * sagelap (~sage@125.sub-70-197-68.myvzw.com) Quit (Ping timeout: 480 seconds)
[20:08] * Cube (~Cube@12.248.40.138) has joined #ceph
[20:16] * xarses (~andreww@12.164.168.115) Quit (Read error: No route to host)
[20:16] * xarses (~andreww@12.164.168.115) has joined #ceph
[20:29] * xarses_ (~andreww@12.164.168.115) has joined #ceph
[20:29] * xarses (~andreww@12.164.168.115) Quit (Read error: Connection reset by peer)
[20:30] <Discard> I have a problem with RGW Logs are here : http://pastebin.com/GpuMZfuz put this file on your chrome: http://p1.13h.com/swift/v1/test/big_buck_bunny.mp4
[20:42] * dpippenger (~riven@cpe-76-166-208-83.socal.res.rr.com) Quit (Quit: Leaving.)
[20:44] * nwat (~textual@c-50-131-197-174.hsd1.ca.comcast.net) Quit (Quit: My MacBook has gone to sleep. ZZZzzz???)
[20:55] * jpds_ is now known as jpds
[20:55] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) has joined #ceph
[21:25] * psomas (~psomas@inferno.cc.ece.ntua.gr) Quit (Ping timeout: 480 seconds)
[21:29] * psomas (~psomas@inferno.cc.ece.ntua.gr) has joined #ceph
[21:44] * Cube1 (~Cube@12.248.40.138) has joined #ceph
[21:44] * Cube (~Cube@12.248.40.138) Quit (Read error: Connection reset by peer)
[21:49] * sarob (~sarob@nat-dip33-wl-g.cfw-a-gci.corp.yahoo.com) has joined #ceph
[21:50] * sarob (~sarob@nat-dip33-wl-g.cfw-a-gci.corp.yahoo.com) Quit (Remote host closed the connection)
[21:50] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) has joined #ceph
[21:53] * diegows (~diegows@200.68.116.185) Quit (Ping timeout: 480 seconds)
[22:01] * diegows (~diegows@200.68.116.185) has joined #ceph
[22:02] * bdonnahue (~tschneide@ool-18bda2d8.dyn.optonline.net) has joined #ceph
[22:03] <bdonnahue> i logged in today to see that my ceph cluster is not healthy
[22:03] <bdonnahue> monclient(hunting): ERROR: missing keyring, cannot use cephx for authentication
[22:03] <bdonnahue> how can i fix this
[22:07] * xmltok (~xmltok@cpe-76-90-130-65.socal.res.rr.com) has joined #ceph
[22:07] * xmltok_ (~xmltok@cpe-76-90-130-65.socal.res.rr.com) Quit (Read error: Connection reset by peer)
[22:08] * xarses_ (~andreww@12.164.168.115) Quit (Ping timeout: 480 seconds)
[22:08] * simulx2 (~simulx@vpn.expressionanalysis.com) has joined #ceph
[22:09] * mnash_ (~chatzilla@66-194-114-178.static.twtelecom.net) has joined #ceph
[22:09] * t0rn (~ssullivan@2607:fad0:32:a02:d227:88ff:fe02:9896) Quit (Quit: Leaving.)
[22:12] <Discard> bdonnahue: it seems that you haven't /etc/ceph/<Cluster_Name>.client.admin.keyring
[22:13] <bdonnahue> ok i see that file is missing on one of my osd's
[22:13] <bdonnahue> can i simply copy this over form another one?
[22:13] * simulx (~simulx@vpn.expressionanalysis.com) Quit (Ping timeout: 480 seconds)
[22:14] * mnash (~chatzilla@vpn.expressionanalysis.com) Quit (Ping timeout: 480 seconds)
[22:14] * mnash_ is now known as mnash
[22:15] <Discard> yep
[22:15] <bdonnahue> im curious why the file might have been deleted
[22:15] <Discard> bdonnahue: it must me the same on all your nodes
[22:15] <bdonnahue> my ceph was health ok when i deployed yesterday
[22:16] <Discard> you could use ceph-deploy admin <node> to push admin key on nodes you want to manage from
[22:17] * Cube1 (~Cube@12.248.40.138) Quit (Quit: Leaving.)
[22:17] * Cube (~Cube@12.248.40.138) has joined #ceph
[22:19] * xarses_ (~andreww@12.164.168.115) has joined #ceph
[22:20] <bdonnahue> copied the keyring and now im seeing the following error:
[22:20] <bdonnahue> 2014-01-03 11:03:06.995116 7f22d08a6700 0 -- :/1001687 >> 10.10.4.5:6789/0 pipe(0x7f22cc024440 sd=3 :0 s=1 pgs=0 cs=0 l=1 c=0x7f22cc0246a0).fault
[22:21] <Discard> when ?
[22:22] <bdonnahue> when what?
[22:22] <pmatulis> who?
[22:22] <bdonnahue> LOL
[22:22] <pmatulis> ;)
[22:22] * Pedras (~Adium@216.207.42.132) has joined #ceph
[22:23] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) Quit (Quit: Computer has gone to sleep.)
[22:25] <bdonnahue> hmm the ceph service on the monitor stopped... strange
[22:30] * jesus (~jesus@emp048-51.eduroam.uu.se) Quit (Ping timeout: 480 seconds)
[22:30] * psomas (~psomas@inferno.cc.ece.ntua.gr) Quit (Ping timeout: 480 seconds)
[22:32] * BillK (~BillK-OFT@106-69-25-13.dyn.iinet.net.au) has joined #ceph
[22:37] * sagelap1 (~sage@38.122.20.226) Quit (Ping timeout: 480 seconds)
[22:38] * rudolfsteiner (~federicon@190.188.67.191) has joined #ceph
[22:38] * diegows (~diegows@200.68.116.185) Quit (Ping timeout: 480 seconds)
[22:40] * sagelap (~sage@38.122.20.226) has joined #ceph
[22:48] * doppelgrau (~doppelgra@pd956d116.dip0.t-ipconnect.de) Quit (Quit: doppelgrau)
[22:48] * jesus (~jesus@emp048-51.eduroam.uu.se) has joined #ceph
[22:48] * diegows (~diegows@200.68.116.185) has joined #ceph
[22:51] <dmsimard> flip some tables
[22:51] <kraken> (??????????????????? ?????????
[22:51] * Cube (~Cube@12.248.40.138) Quit (Quit: Leaving.)
[22:51] * Cube (~Cube@12.248.40.138) has joined #ceph
[22:51] <bdonnahue> thats hilarious... almost choked
[22:52] <bdonnahue> im trying to understand what pools exist in my ceph cluster
[22:53] <bdonnahue> i type ceph osd lspool and see the following:
[22:53] <bdonnahue> 0 data,1 metadata,2 rbd,
[22:53] <bdonnahue> im not sure how to interpret that
[22:54] <bdonnahue> i understand those are the default pools but im not sure what the numbers mean. is it just order?
[23:06] * diegows (~diegows@200.68.116.185) Quit (Ping timeout: 480 seconds)
[23:09] * andreww (~andreww@12.164.168.115) has joined #ceph
[23:09] * xarses_ (~andreww@12.164.168.115) Quit (Quit: Leaving)
[23:15] * diegows (~diegows@200.68.116.185) has joined #ceph
[23:15] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) Quit (Remote host closed the connection)
[23:15] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) has joined #ceph
[23:16] <Pedras> bdonnahue: those are just IDs, as far as I see it
[23:22] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) has joined #ceph
[23:22] <bdonnahue> im trying to figure out how ceph stores data. i realize objects point to files which are distributed accross osds
[23:23] <bdonnahue> but what is a pool? i figured a rbd was an object and thus a file? so if its in a pool is the pool a dir or just meta data?
[23:23] * sarob (~sarob@nat-dip4.cfw-a-gci.corp.yahoo.com) Quit (Ping timeout: 480 seconds)
[23:24] <dmick> bdonnahue: a pool is a grouping of objects. You can have differing access rules per pool, and different crush rules/replication factors
[23:25] <dmick> so basically just a grouping of objects. objects are not files in the traditional sense, but they're basically bags of bits.
[23:25] <dmick> rbd images are made up of multiple objects so that they can be replicated in pieces and spread out over the cluster for redundancy and performance
[23:26] <bdonnahue> ok i thinkk i understand
[23:26] <dmick> you can kind of think of a pool as one level of 'directory', but it implies more, and there's only that level (the object space isn't a hierarchy)
[23:26] * Rocky (~r.nap@188.205.52.204) has joined #ceph
[23:26] * Cube (~Cube@12.248.40.138) Quit (Quit: Leaving.)
[23:30] <bdonnahue> im trying to figure out what rbd's exist
[23:30] <bdonnahue> my ceph health command says health ok
[23:31] <bdonnahue> when i type ceph rbd -p data ls i get an error message about a broken pipe
[23:33] * allsystemsarego (~allsystem@5-12-241-225.residential.rdsnet.ro) Quit (Quit: Leaving)
[23:34] <bdonnahue> i think this could be related to this issue
[23:34] <bdonnahue> http://tracker.ceph.com/issues/6693
[23:36] <dmick> rbds are, by default, in pool rbd
[23:36] <dmick> and the rbd command is not a subcommand of ceph
[23:36] * mattt_ (~textual@cpc25-rdng20-2-0-cust162.15-3.cable.virginm.net) Quit (Quit: Computer has gone to sleep.)
[23:38] <bdonnahue> my monitor keeps crashing. not sure why.
[23:40] <bdonnahue> nothing in ceph.log
[23:43] <bdonnahue> fount the monitor log
[23:44] <bdonnahue> afk thanks everyone time to pass out
[23:45] * mnash_ (~chatzilla@66-194-114-178.static.twtelecom.net) has joined #ceph
[23:51] * mnash (~chatzilla@66-194-114-178.static.twtelecom.net) Quit (Ping timeout: 480 seconds)
[23:51] * mnash_ is now known as mnash

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