#ceph IRC Log

Index

IRC Log for 2016-01-22

Timestamps are in GMT/BST.

[12:24] -dacia.oftc.net- *** Looking up your hostname...
[12:24] -dacia.oftc.net- *** Checking Ident
[12:24] -dacia.oftc.net- *** Found your hostname
[12:25] -dacia.oftc.net- *** No Ident response
[12:25] * CephLogBot (~PircBot@rockbox.widodh.nl) has joined #ceph
[12:25] * Topic is 'http://ceph.com/get || dev channel #ceph-devel || test lab channel #sepia'
[12:25] * Set by scuttlemonkey!~scuttle@nat-pool-rdu-t.redhat.com on Mon Sep 07 00:44:10 CEST 2015
[12:25] * wido__ (~wido@2a00:f10:121:100:4a5:76ff:fe00:199) Quit (Read error: No route to host)
[12:31] * test-cnj (~test_cnj@46.218.86.34) has joined #ceph
[12:33] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[12:35] <MrBy> Heebie: 10.20.30.40/24 and 10.20.30.50/24 are the same subnets. .40 and .50 are just IP's in this subnet. Maybe this was a typo. On the other hand to best guess calculation with as much spare IP's if possible. This isn't really a ceph question, rather then how you design your network
[12:37] * dscastro (~dscastro@190.18.48.177) Quit (Remote host closed the connection)
[12:37] * test-cnj (~test_cnj@46.218.86.34) Quit (Remote host closed the connection)
[12:41] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[12:42] * test-cnj (~test_cnj@46.218.86.34) has joined #ceph
[12:46] * kanagaraj (~kanagaraj@121.244.87.117) Quit (Quit: Leaving)
[12:47] * karnan (~karnan@121.244.87.117) Quit (Remote host closed the connection)
[12:47] * cooldharma06 (~chatzilla@14.139.180.40) has joined #ceph
[12:49] * puvo (~Kristophe@76GAABNQO.tor-irc.dnsbl.oftc.net) has joined #ceph
[12:50] <peeejayz> Anyone got a sure fire way to repair incomplete placement groups? I had a dodgy disk, its now down and out of the cluster but the incompletes are still there.
[12:52] * steveeJ (~junky@141.37.31.187) has joined #ceph
[12:52] <Heebie> MrBy: I meant 10.20.30/0/24 and 20.30.40.0/24, sorry.
[12:53] <Heebie> It's a CEPH question, as far as the settings for public and cluster networks. I need to know if the setting set when the CLUSTER is created has to encompass anything that will ever be used, or if different sets of OSD's (say in different cabinets, suites, data centres etc..) can have their OWN settings for public & cluster networks.
[12:54] * DV (~veillard@2001:41d0:a:f29f::1) Quit (Ping timeout: 480 seconds)
[12:55] <IcePic> Heebie: I think they only need ip access, but if you make the subnet too small, you'd need a router or something in between when adding a new network, so you would get some nodes one hop away, whereas the others would not need that
[12:56] <IcePic> if you are using 10.x or so, just make it big enough for the foreseeable future from the start and go with that.
[12:56] <shinobu> that's true
[12:58] <Heebie> IcePic: So use something like a /16 as a flat network, rather than using a /24? That's what I was thinking. (65,536 nodes should be enough to get me to the point where I can transition to IPv6.. 250 might not be)
[13:00] <MrBy> as far as i understand, they don't do any broadcast stuff, where layer-2 connectivity is needed. thus, they don't have to belong to the same subnet. it is enough if the can connect each other via IP (routing)
[13:01] <IcePic> Heebie: yeah
[13:01] <ojexxi> that may be so, but it sure acts like it
[13:02] * test-cnj (~test_cnj@46.218.86.34) Quit (Remote host closed the connection)
[13:02] <ojexxi> i have an admin server as a vm and even though its just there for ceph-deploy, when i push data to my ceph cluster i get 2+gbit inboud to vmware
[13:03] <ojexxi> when i had my admin server on 1g instead of 10g as a test it smashed the link from the 10g to the 1g switch
[13:03] <ojexxi> however i could not tell you what it is doing
[13:04] <ojexxi> also, i got cephx issues and/or monitors that wouldnt quorum if i didnt have the public network as the nic which had the gateway ip on it, even though mons shouldn't be routing to each other
[13:05] <ojexxi> so it must be broadcasting something
[13:05] <ojexxi> unless its equipment config issue
[13:05] * zhaochao (~zhaochao@125.39.9.144) Quit (Quit: ChatZilla 0.9.92 [Iceweasel 43.0.4/20160107230306])
[13:08] * i_m (~ivan.miro@deibp9eh1--blueice4n0.emea.ibm.com) has joined #ceph
[13:09] * dscastro (~dscastro@host6.181-14-124.telecom.net.ar) has joined #ceph
[13:12] <InIMoeK> swami2 sorry yes
[13:12] <InIMoeK> I wasnt paying attention to irc :)
[13:14] <InIMoeK> ojexxi what data do you push from your admin cluster?
[13:15] <InIMoeK> when using the deph-deploy it fires stuff remotely right? Or are you distributing the packages from the admin node?
[13:15] <ojexxi> its purely got a folder for ceph-deploy, thats it
[13:15] <ojexxi> i only built it because it was part of the concept doc i read
[13:15] <ojexxi> so its doing nada
[13:16] * sleinen (~Adium@2001:620:0:2d:a65e:60ff:fedb:f305) Quit (Read error: Connection reset by peer)
[13:16] <ojexxi> i havent been able to do a test with it off, but from what i can understand, it doesnt participate in general operation
[13:16] * sleinen (~Adium@2001:620:0:82::101) has joined #ceph
[13:16] <InIMoeK> yes that's what I thought
[13:17] <InIMoeK> it's only for management operations because it has the client.admin keyring
[13:18] <InIMoeK> I'm pretty new to Ceph tbh.. so correct me if I'm wrong
[13:18] <InIMoeK> have been reading docs for 3 weeks now ;)
[13:19] * puvo (~Kristophe@76GAABNQO.tor-irc.dnsbl.oftc.net) Quit ()
[13:23] * Mousey (~mollstam@185.60.144.31) has joined #ceph
[13:42] * cooldharma06 (~chatzilla@14.139.180.40) Quit (Quit: ChatZilla 0.9.92 [Iceweasel 21.0/20130515140136])
[13:43] <TheSov-mobile> davidj, yes!
[13:43] * jordanP (~jordan@204.13-14-84.ripe.coltfrance.com) Quit (Ping timeout: 480 seconds)
[13:46] * T1w (~jens@node3.survey-it.dk) Quit (Remote host closed the connection)
[13:53] * nardial (~ls@dslb-178-006-189-150.178.006.pools.vodafone-ip.de) has joined #ceph
[13:53] * Mousey (~mollstam@7V7AAB75J.tor-irc.dnsbl.oftc.net) Quit ()
[13:55] * quinoa (~quinoa@24-148-81-106.c3-0.grn-ubr2.chi-grn.il.cable.rcn.com) has joined #ceph
[13:55] * test-cnj (~test_cnj@46.218.86.34) has joined #ceph
[13:55] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[13:55] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[13:56] <RenoRainz> Hi guys
[13:57] <RenoRainz> Does anyone try to use dmcrypt option with ceph-deploy command ?
[13:57] * allaok (~allaok@machine107.orange-labs.com) has joined #ceph
[14:01] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Remote host closed the connection)
[14:05] * allaok (~allaok@machine107.orange-labs.com) has left #ceph
[14:05] * RenoRainz (~textual@194.73.236.244) Quit (Read error: Connection reset by peer)
[14:07] * storage (~Unforgive@178.162.216.42) has joined #ceph
[14:08] * ade (~abradshaw@dslb-088-075-189-223.088.075.pools.vodafone-ip.de) has joined #ceph
[14:15] * vbellur (~vijay@c-24-62-102-75.hsd1.ma.comcast.net) has joined #ceph
[14:18] <swami2> InIMoeK: thanks...
[14:18] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[14:18] * whaddock (~whaddock@68-185-252-255.dhcp.leds.al.charter.com) Quit (Remote host closed the connection)
[14:24] * georgem (~Adium@104-222-121-229.cpe.teksavvy.com) has joined #ceph
[14:25] <swami2> quick Q: how to take a backup of a pool (reason - doing some changes on cluster, if some thing happens, I can use the backedup pool)
[14:26] * georgem (~Adium@104-222-121-229.cpe.teksavvy.com) Quit ()
[14:28] * danieagle (~Daniel@177.9.76.119) has joined #ceph
[14:33] * vikhyat (~vumrao@121.244.87.116) Quit (Quit: Leaving)
[14:35] * dyasny (~dyasny@cable-192.222.131.135.electronicbox.net) has joined #ceph
[14:35] * T1w (~jens@node3.survey-it.dk) has joined #ceph
[14:36] * jordanP (~jordan@91.199.242.130) has joined #ceph
[14:37] * storage (~Unforgive@76GAABNS5.tor-irc.dnsbl.oftc.net) Quit ()
[14:39] * test-cnj (~test_cnj@46.218.86.34) Quit (Remote host closed the connection)
[14:39] <InIMoeK> swami2 idk.. Hopefully someone else dos
[14:39] <InIMoeK> does
[14:41] <swami2> InIMoeK: ok....
[14:41] * test-cnj (~test_cnj@46.218.86.34) has joined #ceph
[14:41] <swami2> plan to use the cppool
[14:45] * yanzheng2 (~zhyan@182.139.206.224) Quit (Read error: Connection timed out)
[14:45] * yanzheng2 (~zhyan@182.139.206.224) has joined #ceph
[14:49] * xarses_ (~xarses@50.141.35.74) has joined #ceph
[14:50] <peeejayz> Anyone know how I can fit in complete PGs after a failed drive?
[14:53] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[14:55] * andreww (~xarses@50.141.32.222) Quit (Ping timeout: 480 seconds)
[14:56] * mhack (~mhack@66-168-117-78.dhcp.oxfr.ma.charter.com) has joined #ceph
[15:01] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Remote host closed the connection)
[15:02] * RenoRainz (~textual@194.73.236.244) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[15:04] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[15:04] * sleinen1 (~Adium@2001:620:0:2d:a65e:60ff:fedb:f305) has joined #ceph
[15:04] * RenoRainz (~textual@194.73.236.244) Quit ()
[15:05] * sleinen (~Adium@2001:620:0:82::101) Quit (Read error: Connection reset by peer)
[15:07] <mnaser> Would anyone know where I can find the "rados" CLI source code? I'm trying to do something with the Python bindings that's taking 36s which happens in ~1s in the rados CLI
[15:08] * garphy is now known as garphy`aw
[15:08] * overclk (~vshankar@121.244.87.117) Quit (Quit: Zzzz...)
[15:09] * Darius_ is now known as Daznis
[15:10] * brad_mssw (~brad@66.129.88.50) has joined #ceph
[15:10] * test-cnj (~test_cnj@46.218.86.34) Quit (Remote host closed the connection)
[15:16] <InIMoeK> peeejayz when a drive fails the the osd is unreachable
[15:17] <InIMoeK> whenever the OSD is back it should resync?
[15:20] * ulrick40 (~gambit@2a00:1508:1:f010:8055:4f0c:6a97:6e80) has joined #ceph
[15:21] <ulrick40> Hi, I need some websites where i can find news about technology, internet,social network and "OS world". I nedd a site with a lot of news every hour... do u know some? thanks
[15:22] * dvanders_ (~dvanders@dvanders-pro.cern.ch) Quit (Ping timeout: 480 seconds)
[15:24] <IcePic> mnaser: sounds like resolving a hostname-issue
[15:25] <mnaser> IcePic: its not, i've profiled it
[15:25] <mnaser> iterators are slow in python, iterating over 80,000 items is slow
[15:25] * bene_in_mtg (~bene@2601:18c:8501:25e4:ea2a:eaff:fe08:3c7a) has joined #ceph
[15:25] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[15:25] * test-cnj (~test_cnj@46.218.86.34) has joined #ceph
[15:25] * T1w (~jens@node3.survey-it.dk) Quit (Ping timeout: 480 seconds)
[15:25] <nass5> Hi, does anyone know how to dump all parameters of a pool ?
[15:26] <nass5> I know about the "ceph osd pool get <poolname> <param>" but there are more params than the one writtent in the docs...
[15:26] * Skyrider (~SinZ|offl@37.48.81.27) has joined #ceph
[15:27] <Be-El> nass5: 'ceph osd pool ls detail'
[15:27] <nass5> Good one Be-El ! Thanks
[15:32] * mattch (~mattch@w3047.see.ed.ac.uk) Quit (Quit: Leaving.)
[15:33] * mattch (~mattch@w3047.see.ed.ac.uk) has joined #ceph
[15:34] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[15:34] * mattch (~mattch@w3047.see.ed.ac.uk) Quit ()
[15:37] * yanzheng2 (~zhyan@182.139.206.224) Quit (Quit: This computer has gone to sleep)
[15:37] * RenoRainz (~textual@194.73.236.244) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[15:37] * georgem (~Adium@104-222-121-229.cpe.teksavvy.com) has joined #ceph
[15:37] * ulrick40 (~gambit@2a00:1508:1:f010:8055:4f0c:6a97:6e80) Quit (Quit: Sto andando via)
[15:38] * lcurtis (~lcurtis@47.19.105.250) has joined #ceph
[15:39] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[15:42] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[15:42] * RenoRainz (~textual@194.73.236.244) Quit (Read error: Connection reset by peer)
[15:44] * swami2 (~swami@49.32.0.243) Quit (Quit: Leaving.)
[15:45] * rwheeler (~rwheeler@nat-pool-bos-t.redhat.com) has joined #ceph
[15:47] * quinoa (~quinoa@24-148-81-106.c3-0.grn-ubr2.chi-grn.il.cable.rcn.com) Quit (Remote host closed the connection)
[15:48] <peeejayz> InIMoeK: It didn't I found in the documentation to drop my replication min_size to 1 and then it started to repair. I'm guessing that my max_size = 2 so it will always bring everything upto 2 replicants but will cope on 1 until its repaired
[15:49] * harold (~hamiller@71-94-227-123.dhcp.mdfd.or.charter.com) has joined #ceph
[15:49] * harold (~hamiller@71-94-227-123.dhcp.mdfd.or.charter.com) Quit ()
[15:51] * Be-El (~quassel@fb08-bcf-pc01.computational.bio.uni-giessen.de) Quit (Remote host closed the connection)
[15:56] * Skyrider (~SinZ|offl@4MJAABII8.tor-irc.dnsbl.oftc.net) Quit ()
[15:57] * erhudy (uid89730@id-89730.ealing.irccloud.com) has joined #ceph
[16:00] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[16:01] * xarses_ (~xarses@50.141.35.74) Quit (Ping timeout: 480 seconds)
[16:03] * MannerMan (~oscar@user170.217-10-117.netatonce.net) Quit (Remote host closed the connection)
[16:15] * i_m (~ivan.miro@deibp9eh1--blueice4n0.emea.ibm.com) Quit (Read error: Connection reset by peer)
[16:16] * RenoRainz (~textual@194.73.236.244) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[16:16] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[16:17] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[16:20] * fcape (~fcape@107-220-57-73.lightspeed.austtx.sbcglobal.net) Quit (Ping timeout: 480 seconds)
[16:24] * sleinen1 (~Adium@2001:620:0:2d:a65e:60ff:fedb:f305) Quit (Quit: Leaving.)
[16:25] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[16:26] <devicenull> hm, HEALTH_WARN 1 cache pools are missing hit_sets
[16:27] <devicenull> not really sure what how I'm supposed to address that
[16:31] <devicenull> nm, apparently I never configured a hit_set_type
[16:33] * sleinen (~Adium@2001:620:0:2d:a65e:60ff:fedb:f305) has joined #ceph
[16:39] * shaunm (~shaunm@208.102.161.229) Quit (Ping timeout: 480 seconds)
[16:39] * xarses_ (~xarses@64.124.158.100) has joined #ceph
[16:45] * herrsergio (~herrsergi@200.77.224.239) has joined #ceph
[16:46] * herrsergio is now known as Guest470
[16:56] * mattbenjamin (~mbenjamin@aa2.linuxbox.com) has joined #ceph
[16:58] * georgem (~Adium@104-222-121-229.cpe.teksavvy.com) Quit (Quit: Leaving.)
[16:58] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[16:59] * nardial (~ls@dslb-178-006-189-150.178.006.pools.vodafone-ip.de) Quit (Quit: Leaving)
[17:01] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Remote host closed the connection)
[17:01] * bara (~bara@ip4-83-240-10-82.cust.nbox.cz) Quit (Quit: Bye guys! (??????????????????? ?????????)
[17:01] * whaddock (~whaddock@164.111.224.26) has joined #ceph
[17:04] * analbeard (~shw@support.memset.com) Quit (Quit: Leaving.)
[17:04] * dgurtner (~dgurtner@178.197.231.101) Quit (Ping timeout: 480 seconds)
[17:06] * Wielebny (~Icedove@cl-927.waw-01.pl.sixxs.net) Quit (Remote host closed the connection)
[17:07] * Wielebny (~Icedove@cl-927.waw-01.pl.sixxs.net) has joined #ceph
[17:11] * test-cnj (~test_cnj@46.218.86.34) Quit ()
[17:12] * quinoa (~quinoa@24-148-81-106.c3-0.grn-ubr2.chi-grn.il.cable.rcn.com) has joined #ceph
[17:12] * shylesh (~shylesh@59.95.71.124) has joined #ceph
[17:12] * shylesh (~shylesh@59.95.71.124) Quit (autokilled: This host may be infected. Mail support@oftc.net with questions. BOPM (2016-01-22 16:12:23))
[17:13] * xolotl (~PcJamesy@65.19.167.131) has joined #ceph
[17:15] <devicenull> I've got a bunch of pgs stuck stale (an entire pools worth). as faras I can tell, both the public and private networks are fine. enabling debug osd shows a lot of spam of 'share_map_peer 0x768c9ce0 already has epoch 335935'
[17:15] <devicenull> but nothing that really looks like an error
[17:17] <devicenull> can't query any of the pgs stuck, it just hangs
[17:29] * fcape (~fcape@rrcs-97-77-228-30.sw.biz.rr.com) has joined #ceph
[17:34] * bvi (~bastiaan@185.56.32.1) Quit (Ping timeout: 480 seconds)
[17:35] * gregmark (~Adium@68.87.42.115) has joined #ceph
[17:36] * MACscr|lappy (~MACscrlap@173.249.72.254) has joined #ceph
[17:37] * enax (~enax@hq.ezit.hu) Quit (Ping timeout: 480 seconds)
[17:39] * shaunm (~shaunm@cpe-74-132-70-216.kya.res.rr.com) has joined #ceph
[17:41] * joshd1 (~jdurgin@68-119-140-18.dhcp.ahvl.nc.charter.com) has joined #ceph
[17:43] <RenoRainz> hi there
[17:43] <InIMoeK> what does the ceph -s say?
[17:43] * xolotl (~PcJamesy@4MJAABINE.tor-irc.dnsbl.oftc.net) Quit ()
[17:43] <InIMoeK> devicenull
[17:43] <RenoRainz> I try to setup osd with dmcrypt and I'm facing an issue
[17:43] <devicenull> dunno, I changed the broken pool over to another crush ruleset
[17:43] <devicenull> and it's rebuilding now, not sure wtf happened
[17:44] * dugravot6 (~dugravot6@dn-infra-04.lionnois.univ-lorraine.fr) Quit (Remote host closed the connection)
[17:44] <InIMoeK> strange
[17:44] <RenoRainz> I prepared a disk with this command : ceph-deploy osd prepare ceph-osd-1:/dev/sdc --dmcrypt --dmcrypt-key-dir /root/keydir
[17:44] <RenoRainz> it's ok
[17:45] <RenoRainz> but when I try to activate it fail
[17:45] <InIMoeK> I just simulated a disk failure
[17:45] * fcape (~fcape@rrcs-97-77-228-30.sw.biz.rr.com) Quit (Quit: leaving)
[17:45] * fcape (~fcape@rrcs-97-77-228-30.sw.biz.rr.com) has joined #ceph
[17:45] <RenoRainz> [ceph-osd-1][WARNIN] mount: unknown filesystem type 'crypto_LUKS' / [ceph-osd-1][WARNIN] ceph-disk: Mounting filesystem failed: Command '['/bin/mount', '-t', 'crypto_LUKS', '-o', '', '--', '/dev/sdc1', '/var/lib/ceph/tmp/mnt.C0wSgD']' returned non-zero exit status 32
[17:45] * zaitcev (~zaitcev@c-50-130-189-82.hsd1.nm.comcast.net) has joined #ceph
[17:45] <RenoRainz> Do you guys have any ideas ?
[17:46] * daviddcc (~dcasier@84.197.151.77.rev.sfr.net) has joined #ceph
[17:46] * dscastro_ (~dscastro@host6.181-14-124.telecom.net.ar) has joined #ceph
[17:46] * dscastro (~dscastro@host6.181-14-124.telecom.net.ar) Quit (Read error: Connection reset by peer)
[17:47] * reed (~reed@75-101-54-18.dsl.static.fusionbroadband.com) has joined #ceph
[17:47] * derjohn_mobi (~aj@2001:6f8:1337:0:c1dd:c388:a6f3:c9d6) Quit (Ping timeout: 480 seconds)
[17:48] <InIMoeK> I pulled a disk, and reseated it after 5 min
[17:49] <InIMoeK> XFS was like : "Oh look, it's a duplicate UUID"
[17:49] <InIMoeK> lets mount it under a completly different drive letter
[17:49] <InIMoeK> DERP
[17:49] <TheSov> u cloned a disk?
[17:50] <InIMoeK> but the most genius thing about it is that the reboot of the OSD node fixed the problem haha
[17:50] <InIMoeK> "have you tried turning it off and on again"
[17:50] <InIMoeK> TheSov no reseated the same disk
[17:50] <TheSov> lol
[17:51] <InIMoeK> here wait
[17:51] <TheSov> i guess the pipe had not yet broken on you yanking it
[17:52] * georgem (~Adium@184.151.179.52) has joined #ceph
[17:53] <InIMoeK> Jan 22 15:51:19 ceph04 kernel: [681705.030585] XFS (sdw1): Filesystem has duplicate UUID 94f0e76f-9c7d-4b3b-b06a-5582cdc4153d - can't mount
[17:53] <InIMoeK> Jan 22 15:51:19 ceph04 kernel: [681705.253516] XFS (sdw1): Filesystem has duplicate UUID 94f0e76f-9c7d-4b3b-b06a-5582cdc4153d - can't mount
[17:53] <InIMoeK> Jan 22 15:51:21 ceph04 kernel: [681706.622865] XFS (sdl1): metadata I/O error: block 0x39e8338 ("xfs_buf_iodone_callbacks") error 19 numblks 8
[17:53] <InIMoeK> Jan 22 15:51:26 ceph04 kernel: [681711.638668] XFS (sdl1): metadata I/O error: block 0x39e8338 ("xfs_buf_iodone_callbacks") error 19 numblks 8
[17:53] <InIMoeK> sdl is the old mount point
[17:53] <InIMoeK> sdw was XFS's great idea
[17:53] <TheSov> lol
[17:54] <TheSov> so just an FYI for the future, its best to down an osd before reseating it LOL
[17:56] <InIMoeK> hehe true
[17:56] <InIMoeK> but what if the device failed
[17:56] <InIMoeK> then it's the same situation
[17:56] * ZombieL (~TGF@nl8x.mullvad.net) has joined #ceph
[17:59] <rkeene> Tell the kernel to delete sdl
[17:59] * Discovery (~Discovery@178.239.49.70) has joined #ceph
[18:00] * quinoa (~quinoa@24-148-81-106.c3-0.grn-ubr2.chi-grn.il.cable.rcn.com) Quit (Remote host closed the connection)
[18:00] <rkeene> i.e., echo 1 > /sys/block/sdl/device/delete
[18:00] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[18:00] <TheSov> echo 1 > /sys/block/SDX/device/delete
[18:00] <TheSov> damnit
[18:00] <TheSov> u beat me
[18:01] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Remote host closed the connection)
[18:01] <rkeene> Well, you are slow...
[18:01] * lxo (~aoliva@lxo.user.oftc.net) Quit (Ping timeout: 480 seconds)
[18:02] <rkeene> :-)
[18:02] <TheSov> lol
[18:02] * georgem (~Adium@184.151.179.52) Quit (Quit: Leaving.)
[18:03] * blizzow (~jburns@50.243.148.102) has joined #ceph
[18:04] * georgem (~Adium@206.108.127.16) has joined #ceph
[18:04] * RenoRainz (~textual@194.73.236.244) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[18:05] <TheSov> what should i do for my next ceph video?
[18:06] <TheSov> FC?
[18:06] <TheSov> that might be hard to do on screen lol
[18:06] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[18:10] * joshd1 (~jdurgin@68-119-140-18.dhcp.ahvl.nc.charter.com) Quit (Quit: Leaving.)
[18:10] <blizzow> Where does the cephlogbot log to?
[18:13] <TheSov> blizzow, IN A ZOO!
[18:13] <TheSov> HES AN ANIMAL!
[18:13] * reed (~reed@75-101-54-18.dsl.static.fusionbroadband.com) Quit (Quit: Ex-Chat)
[18:14] * bitserker (~toni@88.87.194.130) Quit (Ping timeout: 480 seconds)
[18:15] * wjw-freebsd (~wjw@vpn.ecoracks.nl) Quit (Ping timeout: 480 seconds)
[18:17] <stupidnic> does anybody have a method for automatically mounting bcache devices using udev?
[18:18] * swami1 (~swami@27.7.169.140) has joined #ceph
[18:21] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[18:25] * Geph_ (~Geoffrey@41.77.153.99) has joined #ceph
[18:26] <devicenull> turns out my issue was I restarted the osds, so osd crush location hook was executed again and moved all the OSDs out of the correct place in the tree
[18:26] * ZombieL (~TGF@84ZAABR4D.tor-irc.dnsbl.oftc.net) Quit ()
[18:27] * GeoTracer (~Geoffrey@105.5.73.12) has joined #ceph
[18:29] * topro_ (~tobi@pD9F855EB.dip0.t-ipconnect.de) has joined #ceph
[18:29] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[18:29] <topro_> TheSov: ping
[18:30] * lxo (~aoliva@lxo.user.oftc.net) has joined #ceph
[18:30] <TheSov> ?
[18:30] * vasu (~vasu@c-73-231-60-138.hsd1.ca.comcast.net) has joined #ceph
[18:31] <topro_> we've been talking about cache tiering of cephfs yesterday, remember?
[18:31] * dscastro_ (~dscastro@host6.181-14-124.telecom.net.ar) Quit (Read error: No route to host)
[18:31] <TheSov> yes indeed
[18:31] * whaddock (~whaddock@164.111.224.26) Quit (Remote host closed the connection)
[18:31] * dscastro (~dscastro@host6.181-14-124.telecom.net.ar) has joined #ceph
[18:31] * sleinen (~Adium@2001:620:0:2d:a65e:60ff:fedb:f305) Quit (Quit: Leaving.)
[18:32] <topro_> have been thinking and reading a lot, but seems it won't work due to different ceph limitations. anyway will come back to you later, got an interrupt from my wife ;)
[18:32] * Nats (~natscogs@114.31.195.238) Quit (Read error: Connection reset by peer)
[18:33] * pabluk is now known as pabluk_
[18:33] * Nats (~natscogs@114.31.195.238) has joined #ceph
[18:34] * Geph_ (~Geoffrey@41.77.153.99) Quit (Ping timeout: 480 seconds)
[18:34] * srk (~oftc-webi@32.97.110.56) has joined #ceph
[18:42] * jordanP (~jordan@91.199.242.130) Quit (Quit: Leaving)
[18:44] * thomnico (~thomnico@2a01:e35:8b41:120:9588:ee4a:8471:6231) Quit (Quit: Ex-Chat)
[18:44] <srk> I got one pg stuck at Incomplete.The osds its probing are all up.
[18:44] <TheSov> srk, how much free space you got?
[18:44] <TheSov> and are you on infernalis?
[18:44] <srk> this is Firefly
[18:45] <srk> an old cluster, The cluster is 4 node cluster with 96 osds, ssd journal and 3 copies
[18:46] <srk> about 65% RAW used in the cluster
[18:47] <srk> 2 osds are marked out of the cluster, because of disk issues and they are still out.
[18:47] <srk> but, rebalance is complete and left with 1 incomplete pg
[18:47] <srk> GLOBAL: SIZE AVAIL RAW USED %RAW USED 341T 122T 219T 64.25
[18:52] <topro_> TheSov: re
[18:52] <TheSov> ?
[18:52] <topro_> to the limitations I think there are
[18:53] * w2k (~Hejt@76GAABN5A.tor-irc.dnsbl.oftc.net) has joined #ceph
[18:53] <topro_> gregsfortytwo yesterday told someone else that ceph still can only handle a single fs per cluster, despite the fact that beginning with 0.80 there was a commad added "ceph fs ls" which litst all created filesystems of cluster
[18:54] <topro_> with ceph cache tiering I would need one fs per dc, each consisting of an metadata and data cache-tier, right?
[18:55] <srk> TheSov,any ideas to get rid of incomplete pg?
[18:55] <topro_> then I would have to tell each client the exact fs to connect to, which is also not possible with mount.ceph as of yet I think
[18:55] <topro_> or did I get you wrong?
[18:57] <topro_> srk: http://docs.ceph.com/docs/master/rados/operations/placement-groups/#revert-lost ?
[18:59] * InIMoeK (~InIMoeK@95.170.93.16) Quit (Ping timeout: 480 seconds)
[19:00] * ngoswami (~ngoswami@121.244.87.116) Quit (Quit: Leaving)
[19:00] <srk> Thanks, I was reading it before. Wondering whether there are any alternatives to recover it before marking it Lost :)
[19:00] * ircolle (~ircolle@dhcp-18-189-4-63.dyn.mit.edu) has joined #ceph
[19:00] * ircolle (~ircolle@dhcp-18-189-4-63.dyn.mit.edu) has left #ceph
[19:01] <topro_> srk: don't know if there are alternatives to recover, maybe betrer ask someone else. I only know how to get rid of invalid pgs, thought that is what you were asking for
[19:01] * swami1 (~swami@27.7.169.140) Quit (Quit: Leaving.)
[19:02] * RenoRainz (~textual@194.73.236.244) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[19:03] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[19:03] <topro_> TheSov: to make clear how I would love to implement cache tiering (due to what I learned from you yesterday) would be to have a couple of OSD with standard storage (i.e. erasure coding) on spinners holding normal pools for data and metadata. those pools replicated over machines in three different dc on a campus...
[19:04] * RenoRainz (~textual@194.73.236.244) has joined #ceph
[19:04] <TheSov> right
[19:04] <topro_> then in every DC have one cache-tiering machine providing a cache for metadata and data...
[19:04] * dscastro (~dscastro@host6.181-14-124.telecom.net.ar) Quit (Remote host closed the connection)
[19:04] <topro_> make up for three cahce-tier pools per standard pool, right so far?
[19:05] * dscastro (~dscastro@host6.181-14-124.telecom.net.ar) has joined #ceph
[19:05] <TheSov> im pretty sure its 1 cache pool per normal pool
[19:05] * efirs (~firs@c-50-185-70-125.hsd1.ca.comcast.net) has joined #ceph
[19:05] <topro_> I was afraid you would say something like that :(
[19:06] <topro_> so I don't even come close to where I was expecting the limitations ;)
[19:07] <topro_> as I didn't find a clear indication of only one cache per pool (as of http://docs.ceph.com/docs/master/rados/operations/cache-tiering/) and reading the command "ceph osd tier add ..." I was assuming you can have multiple cache tiers per pool
[19:08] * dvanders (~dvanders@46.227.20.178) has joined #ceph
[19:09] <TheSov> no man sorry
[19:09] <TheSov> its 1 cache per pool
[19:10] * wkennington (~william@c-50-184-242-109.hsd1.ca.comcast.net) Quit (Quit: Bye.)
[19:10] <topro_> but you get the idea of what I intend to do. do you think its crazy or is it just that ceph isnt there yet
[19:10] <TheSov> you can do it if you avoid cephfs and use RBD's with local file servers
[19:11] * wkennington (~william@c-50-184-242-109.hsd1.ca.comcast.net) has joined #ceph
[19:11] * DanFoster (~Daniel@2a00:1ee0:3:1337:5cb3:ce57:9ece:e651) Quit (Quit: Leaving)
[19:11] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[19:12] <topro_> well the point is I love the concepts of cephfs. but thene whenever I mention it I feel like most people try to avoid it the best they can
[19:13] * dscastro (~dscastro@host6.181-14-124.telecom.net.ar) Quit (Ping timeout: 480 seconds)
[19:13] <TheSov> i dont unstand the purpose of hitting local only disk. do you have a bandwidth constraint?
[19:15] <topro_> actually you are right, I don't have a bandwidth problem so maybe I'm looking into the wrong direction. Its just that doing it that way would result in a very neat design. As I mentioned yesterday my real problem is that I only get about 50-80 IOPS on cephfs and I cannot find the real bottleneck. seems neither to be OSD-IOPS, nor CPU, nor Bandwith. Don't get it
[19:17] <topro_> ^^ that when running fio with supplied example config for iometer emulation
[19:17] <TheSov> for cephfs to be fast(er) you need quite a beefy processor on a dedicated system
[19:17] * nils_ (~nils_@doomstreet.collins.kg) Quit (Quit: This computer has gone to sleep)
[19:18] <topro_> each of my machines runs on bare metal 8 core operons, not beefy enough?
[19:19] <topro_> still I see no real cpu load when running that benchmark
[19:22] <topro_> i think its more like IO latency maybe due to the fact that every cephfs IO takes multiple backend IOs, each one added to network latency (which should be too low to do any harm, RTT is ~100us)
[19:23] * w2k (~Hejt@76GAABN5A.tor-irc.dnsbl.oftc.net) Quit ()
[19:24] * GeoTracer (~Geoffrey@105.5.73.12) Quit (Read error: Connection reset by peer)
[19:24] * Geoff (~Geoffrey@41.77.153.99) has joined #ceph
[19:26] * MACscr|lappy (~MACscrlap@173.249.72.254) Quit (Ping timeout: 480 seconds)
[19:27] * angdraug (~angdraug@64.124.158.100) has joined #ceph
[19:28] * NTmatter (uid141693@id-141693.highgate.irccloud.com) has joined #ceph
[19:28] <devicenull> with cache tiering, what does my authentication need to look like?
[19:29] * dvanders (~dvanders@46.227.20.178) Quit (Remote host closed the connection)
[19:29] <TheSov> rwx to both cache tier and backend
[19:29] <devicenull> nothing else?
[19:29] <TheSov> not that i am aware
[19:30] <NTmatter> Might also need mon 'allow r' :)
[19:30] * efirs (~firs@c-50-185-70-125.hsd1.ca.comcast.net) Quit (Read error: No route to host)
[19:30] <TheSov> thats a given :P
[19:30] * efirs (~firs@c-50-185-70-125.hsd1.ca.comcast.net) has joined #ceph
[19:30] <TheSov> mon allow r should be default. i dont get the need for that
[19:31] <TheSov> what situation would you allow rwx to osd and disallow monitor?
[19:31] <TheSov> i am ignorant of that need
[19:31] <NTmatter> errr, that should have been mds 'allow'
[19:31] <devicenull> ya rwx to the cache and backend seems to have worked.. oddly I swear that was causing problems yesterday
[19:32] <TheSov> maybe u did just rw?
[19:32] <devicenull> or maybe puppet reverted my changes behind my back :)
[19:32] * reed (~reed@216.9.110.9) has joined #ceph
[19:33] * kawa2014 (~kawa@195.110.70.45) Quit (Ping timeout: 480 seconds)
[19:33] * Geoff (~Geoffrey@41.77.153.99) Quit (Ping timeout: 480 seconds)
[19:36] * jordanP (~jordan@bdv75-2-81-57-250-57.fbx.proxad.net) has joined #ceph
[19:38] * efirs (~firs@c-50-185-70-125.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[19:39] * georgem (~Adium@206.108.127.16) Quit (Ping timeout: 480 seconds)
[19:45] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[19:46] * Scymex (~rogst@watchme.tor-exit.network) has joined #ceph
[19:49] * ircolle (~ircolle@dhcp-18-189-4-63.dyn.MIT.EDU) has joined #ceph
[19:49] * ircolle (~ircolle@dhcp-18-189-4-63.dyn.MIT.EDU) has left #ceph
[19:50] * jordanP (~jordan@bdv75-2-81-57-250-57.fbx.proxad.net) Quit (Ping timeout: 480 seconds)
[19:53] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[19:55] * dscastro (~dscastro@190.18.48.177) has joined #ceph
[19:56] <TheSov> w00t my ceph install video has almost 500 views! WHAT LOW STANDARDS I HAVE!
[19:57] <NTmatter> Congrats! :)
[19:58] <SamYaple> TheSov: any you passed the point youtube starts validating views which means those are legit and not just you rewatching it!
[19:59] <TheSov> lol
[19:59] <TheSov> i dont think i will ever reach pewdiepie notariaty but if the videos can help anyone build a working ceph cluster that will make me happy
[20:00] <TheSov> i watch other installation or deployment videos and its just music playing in the backround or silence on a potato resolution
[20:00] <TheSov> i was like, im gonna make my own1
[20:02] * reed (~reed@216.9.110.9) Quit (Quit: Ex-Chat)
[20:04] <NTmatter> heh, "The PewdiePie of Storage" :) This is the beginning, folks.
[20:05] * TomasCZ (~TomasCZ@yes.tenlab.net) has joined #ceph
[20:07] * RenoRainz (~textual@194.73.236.244) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[20:09] * kawa2014 (~kawa@82-68-198-196.dsl.in-addr.zen.co.uk) has joined #ceph
[20:09] <TheSov> hell yeah that would own
[20:10] <TheSov> why do people hate LIO?
[20:12] * dgbaley27 (~matt@75.148.118.217) has joined #ceph
[20:14] * quinoa (~quinoa@c-50-165-115-208.hsd1.il.comcast.net) has joined #ceph
[20:16] * Scymex (~rogst@7V7AAB8DT.tor-irc.dnsbl.oftc.net) Quit ()
[20:16] * shaunm (~shaunm@cpe-74-132-70-216.kya.res.rr.com) Quit (Ping timeout: 480 seconds)
[20:16] * kawa2014 (~kawa@82-68-198-196.dsl.in-addr.zen.co.uk) Quit (Quit: Leaving)
[20:21] * debian112 (~bcolbert@24.126.201.64) Quit (Quit: Leaving.)
[20:23] * cpetersen_ (~cpetersen@208.38.1.66) has joined #ceph
[20:27] * rwheeler (~rwheeler@nat-pool-bos-t.redhat.com) Quit (Quit: Leaving)
[20:27] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[20:27] <blizzow> 'cause it's SCSI.
[20:29] * sleinen (~Adium@2001:620:1000:3:a65e:60ff:fedb:f305) has joined #ceph
[20:29] * cpetersen (~cpetersen@208.38.1.66) Quit (Ping timeout: 480 seconds)
[20:35] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[20:36] * debian112 (~bcolbert@24.126.201.64) has joined #ceph
[20:41] * sleinen (~Adium@2001:620:1000:3:a65e:60ff:fedb:f305) Quit (Quit: Leaving.)
[20:42] <dack> TheSov: what's your youtube channel?
[20:42] * dneary (~dneary@nat-pool-bos-u.redhat.com) Quit (Ping timeout: 480 seconds)
[20:42] <TheSov> bovinespongiformflu
[20:42] <TheSov> just google
[20:43] <TheSov> ignore the electronics videos
[20:43] <dack> what if i also like electronics :)
[20:43] <TheSov> well then dont ignore them ?
[20:43] <dack> my other hat is EE
[20:44] <TheSov> well then you might like this, im building an EV!
[20:44] <dack> nice! i'm actually in the middle of building an electric bike right now.
[20:45] <TheSov> oh yeah? well my drive motor is 15 treadmill motors ganged together!
[20:45] * mykola (~Mikolaj@91.225.203.16) has joined #ceph
[20:45] <dack> wow. are they brushed motors?
[20:46] <TheSov> they are but they run at 130 volts, i have a mosfet for each motor. it distributes the load and a treadmill motor is hella cheap
[20:46] * clarjon1 (~AotC@4MJAABIVG.tor-irc.dnsbl.oftc.net) has joined #ceph
[20:46] <TheSov> irfp250's are the drive mosfets
[20:46] <TheSov> fod3181 to drive them
[20:47] <dack> how much current?
[20:47] <TheSov> 30 amps per motor
[20:47] <dack> get some good heatsinking on those irfp250's :)
[20:47] <TheSov> roughly 80 horsepower
[20:48] <TheSov> indeed. i wanna put the heatsink through the hood like a hoodscoop
[20:48] <TheSov> it would look very cool i think
[20:49] <dack> are you building the frame, or using an existing one?
[20:49] <TheSov> im looking for a used car with a bad engine now
[20:49] <TheSov> i even got a ad out on reddit lol
[20:50] <rkeene> For my EV I plan to use a premade engine... my main issue is figuring out where to shove all the batteries :-/
[20:54] <TheSov> nice what kind of battery?
[20:54] <rkeene> It's been a while since I looked, some gelpacks
[20:54] <TheSov> oh i was gonna go for some 38120's
[20:54] <TheSov> their like 18650's but ...bigger
[20:55] * lxo (~aoliva@lxo.user.oftc.net) Quit (Ping timeout: 480 seconds)
[20:56] * InIMoeK (~InIMoeK@029-226-128-083.dynamic.caiway.nl) has joined #ceph
[20:58] * puffy (~puffy@c-50-131-179-74.hsd1.ca.comcast.net) has joined #ceph
[21:02] * daviddcc (~dcasier@84.197.151.77.rev.sfr.net) Quit (Remote host closed the connection)
[21:03] * dgbaley27 (~matt@75.148.118.217) Quit (Quit: Leaving.)
[21:03] <InIMoeK> TheSov a video how to deal with snapshots would be interesting
[21:03] <TheSov> ohhh thats a cool one
[21:03] <TheSov> ill do that next
[21:03] * shaunm (~shaunm@ms-208-102-107-95.gsm.cbwireless.com) has joined #ceph
[21:04] <InIMoeK> awesome!
[21:05] * Geoff (~Geoffrey@169-0-145-75.ip.afrihost.co.za) has joined #ceph
[21:05] <InIMoeK> btw I'll test the reseat of the disk again on monday, and then tell the kernel to forget the disk mount
[21:05] <InIMoeK> because rebooting the whole node is a bit of a hassle :)
[21:05] * gregmark (~Adium@68.87.42.115) Quit (Quit: Leaving.)
[21:08] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[21:09] * RenoRainz (~textual@149.18.54.89) has joined #ceph
[21:09] * Geph_ (~Geoffrey@169-0-145-75.ip.afrihost.co.za) has joined #ceph
[21:11] * puffy (~puffy@c-50-131-179-74.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[21:12] * rotbeard (~redbeard@p54BE7C27.dip0.t-ipconnect.de) has joined #ceph
[21:13] * srk (~oftc-webi@32.97.110.56) Quit (Ping timeout: 480 seconds)
[21:14] * mortn (~mortn@217-215-219-69-no229.tbcn.telia.com) Quit (Remote host closed the connection)
[21:14] * GeoTracer (~Geoffrey@169-0-145-75.ip.afrihost.co.za) has joined #ceph
[21:16] * clarjon1 (~AotC@4MJAABIVG.tor-irc.dnsbl.oftc.net) Quit ()
[21:16] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[21:17] * RenoRainz (~textual@149.18.54.89) Quit (Read error: Connection reset by peer)
[21:18] * RenoRainz (~textual@149.18.54.89) has joined #ceph
[21:22] * Geph_ (~Geoffrey@169-0-145-75.ip.afrihost.co.za) Quit (Quit: Leaving)
[21:22] * branto1 (~borix@ip-78-102-208-28.net.upcbroadband.cz) has left #ceph
[21:22] * GeoTracer (~Geoffrey@169-0-145-75.ip.afrihost.co.za) Quit (Quit: Leaving)
[21:25] * Geoff is now known as Geph
[21:25] * garphy`aw is now known as garphy
[21:26] * ibravo (~ibravo@72.198.142.104) has joined #ceph
[21:27] * dscastro (~dscastro@190.18.48.177) Quit (Remote host closed the connection)
[21:28] * dscastro (~dscastro@190.18.48.177) has joined #ceph
[21:29] * efirs (~firs@c-50-185-70-125.hsd1.ca.comcast.net) has joined #ceph
[21:29] * Crisco (~Misacorp@188.166.80.153) has joined #ceph
[21:32] * derjohn_mob (~aj@x4db004df.dyn.telefonica.de) has joined #ceph
[21:32] * cathode (~cathode@50-198-166-81-static.hfc.comcastbusiness.net) has joined #ceph
[21:33] * Kupo1 (~tyler.wil@23.111.254.159) has joined #ceph
[21:35] * DV (~veillard@2001:41d0:a:f29f::1) has joined #ceph
[21:36] * RenoRainz (~textual@149.18.54.89) Quit (Quit: My Mac has gone to sleep. ZZZzzz???)
[21:36] * dscastro (~dscastro@190.18.48.177) Quit (Ping timeout: 480 seconds)
[21:39] * Kupo2 (~tyler.wil@23.111.254.159) Quit (Ping timeout: 480 seconds)
[21:44] <devicenull> is it possible to tell how much space is available/used for a particular crush root?
[21:45] <devicenull> the overall cluster stats seem to be less useful when you've got multiple roots
[21:46] <devicenull> or laternatively, usuage by pool
[21:46] <devicenull> I guess 'rados df' gives me used size
[21:48] * whaddock (~whaddock@75-138-225-167.dhcp.jcsn.tn.charter.com) has joined #ceph
[21:50] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[21:51] * dneary (~dneary@nat-pool-bos-u.redhat.com) has joined #ceph
[21:54] * mykola (~Mikolaj@91.225.203.16) Quit (Read error: No route to host)
[21:54] <InIMoeK> devicenull rados df -p pool
[21:54] * mykola (~Mikolaj@91.225.203.16) has joined #ceph
[21:54] <InIMoeK> thats nothing really diffrent from the normal df haha
[21:54] <InIMoeK> probably noth the thing you mean
[21:54] <devicenull> InIMoeK: yea, not really
[21:54] <devicenull> I have a pool that's made up of a subset of OSDs
[21:55] <devicenull> and I'd like to track usage of just that pool
[21:55] <devicenull> I can get the total used space pretty easily, but not free space
[21:58] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[21:59] <InIMoeK> hmm
[21:59] <InIMoeK> ceph osd df doesnt show the pools
[21:59] * Crisco (~Misacorp@4MJAABIWI.tor-irc.dnsbl.oftc.net) Quit ()
[22:00] <TheSov> ceph osd lspools
[22:00] <TheSov> jeez people
[22:02] <rkeene> Hmm ?
[22:02] <InIMoeK> lol but that isnt really the answer to his question
[22:03] <InIMoeK> now that I'm reading again I lost it again :D
[22:04] <InIMoeK> everything in ceph is thin provisioning right?
[22:04] * xul (~Chaos_Lla@185.60.144.31) has joined #ceph
[22:05] * whaddock (~whaddock@75-138-225-167.dhcp.jcsn.tn.charter.com) Quit (Remote host closed the connection)
[22:05] * whaddock (~whaddock@164.111.185.69) has joined #ceph
[22:06] <TheSov> you can thick provision them manually
[22:06] <InIMoeK> ah ok
[22:07] <TheSov> and by manually i mean literally overwrite the luns with 1's
[22:07] <TheSov> err rbd's
[22:07] * danieagle (~Daniel@177.9.76.119) Quit (Quit: Obrigado por Tudo! :-) inte+ :-))
[22:08] <InIMoeK> the reason why I'm asking is that devicenull wants to see the free space within the pools
[22:08] <InIMoeK> haha dat SAN guy :D
[22:08] <InIMoeK> luns
[22:08] <TheSov> excuse me I'm a storage admin by trade
[22:08] <blizzow> Anyone here know how a ceph client decides which mon to talk to?
[22:09] <TheSov> blizzow, which ever one it responds first :D
[22:11] <blizzow> It's sending out a multicast request?
[22:12] <InIMoeK> I think just a normal packet since multicast isnt accepted in every network?
[22:12] <InIMoeK> or routed..
[22:14] <devicenull> ceph definitely doesnt use multicast
[22:16] * pabluk_ is now known as pabluk
[22:16] <devicenull> InIMoeK: I can probably work with 'ceph osd df'
[22:16] <devicenull> didn't realize that was a command
[22:17] <InIMoeK> me neither
[22:17] <InIMoeK> but I noticed in documentation ceph osd ls was workign as well
[22:17] <InIMoeK> haha
[22:17] <InIMoeK> so I tried df
[22:18] <InIMoeK> ceph osd df tree
[22:18] <InIMoeK> try that one
[22:18] <InIMoeK> a bit more info
[22:20] * petracvv (~quassel@c-50-158-9-81.hsd1.il.comcast.net) has joined #ceph
[22:21] <devicenull> ah yea, that has exactly what I want
[22:21] <devicenull> usage by root, thanks
[22:21] <InIMoeK> haha yay
[22:21] <InIMoeK> my first bit of help
[22:23] * dneary (~dneary@nat-pool-bos-u.redhat.com) Quit (Ping timeout: 480 seconds)
[22:24] <devicenull> woo it supports --format json as well
[22:24] <devicenull> nice and easy
[22:24] <InIMoeK> InIMoeK ceph osd df tree > /dev/null
[22:24] <InIMoeK> to sum our conversation up
[22:24] <InIMoeK> :D
[22:24] <devicenull> lol
[22:24] <InIMoeK> awesome!
[22:25] * rendar (~I@host63-29-dynamic.247-95-r.retail.telecomitalia.it) Quit (Ping timeout: 480 seconds)
[22:25] * DV (~veillard@2001:41d0:a:f29f::1) Quit (Remote host closed the connection)
[22:26] * shaunm (~shaunm@ms-208-102-107-95.gsm.cbwireless.com) Quit (Ping timeout: 480 seconds)
[22:27] * rendar (~I@host63-29-dynamic.247-95-r.retail.telecomitalia.it) has joined #ceph
[22:27] * whaddock (~whaddock@164.111.185.69) Quit (Ping timeout: 480 seconds)
[22:29] * whaddock (~whaddock@75-138-225-167.dhcp.jcsn.tn.charter.com) has joined #ceph
[22:29] * enax (~enax@94-21-125-51.pool.digikabel.hu) has joined #ceph
[22:30] * whaddock (~whaddock@75-138-225-167.dhcp.jcsn.tn.charter.com) Quit (Remote host closed the connection)
[22:30] * whaddock (~whaddock@164.111.187.71) has joined #ceph
[22:31] <blizzow> I just converted all my libvirt based VMs that were running off local disk qcow files to run off image files in my ceph pool. The first 70% went really well, and ceph performed admirably. Now it seems that large disk IO on one VM will cause sluggish behavior on disk ops across all other VMs. A) is there a way to see per image IOps?
[22:32] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[22:33] <blizzow> I'd like to see if a couple VMs are just being mean to the ceph pool.
[22:33] * DV (~veillard@2001:41d0:a:f29f::1) has joined #ceph
[22:34] <InIMoeK> I think you should check that at your hypervisor level
[22:34] <InIMoeK> since ceph will probably return the aggregated iops
[22:34] * xul (~Chaos_Lla@76GAABOBN.tor-irc.dnsbl.oftc.net) Quit ()
[22:34] <InIMoeK> since it ( what I think ) comes from one pool
[22:36] <InIMoeK> or do you have different pools defined
[22:36] <InIMoeK> BlackDex
[22:36] <InIMoeK> eeh sorry.. blizzow
[22:40] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[22:43] * Guest470 is now known as herrsergio
[22:44] * topro_ (~tobi@pD9F855EB.dip0.t-ipconnect.de) Quit (Quit: Konversation terminated!)
[22:47] * pabluk is now known as pabluk_
[22:52] * ggg (~Mraedis@nooduitgang.schmutzig.org) has joined #ceph
[22:53] * enax (~enax@94-21-125-51.pool.digikabel.hu) Quit (Ping timeout: 480 seconds)
[22:54] * dscastro (~dscastro@190.18.48.177) has joined #ceph
[22:55] * sleinen (~Adium@2001:620:1000:3:a65e:60ff:fedb:f305) has joined #ceph
[23:01] * vasu (~vasu@c-73-231-60-138.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[23:04] * vasu (~vasu@c-73-231-60-138.hsd1.ca.comcast.net) has joined #ceph
[23:05] * bene_in_mtg (~bene@2601:18c:8501:25e4:ea2a:eaff:fe08:3c7a) Quit (Quit: Konversation terminated!)
[23:09] * steveeJ (~junky@141.37.31.187) Quit (Ping timeout: 480 seconds)
[23:11] * brad_mssw (~brad@66.129.88.50) Quit (Quit: Leaving)
[23:11] * InIMoeK (~InIMoeK@029-226-128-083.dynamic.caiway.nl) Quit (Ping timeout: 480 seconds)
[23:14] * dscastro (~dscastro@190.18.48.177) Quit (Remote host closed the connection)
[23:14] * haomaiwang (~haomaiwan@103.15.217.218) has joined #ceph
[23:19] * dscastro (~dscastro@190.18.48.177) has joined #ceph
[23:22] * ggg (~Mraedis@7V7AAB8F3.tor-irc.dnsbl.oftc.net) Quit ()
[23:22] * haomaiwang (~haomaiwan@103.15.217.218) Quit (Ping timeout: 480 seconds)
[23:26] * mykola (~Mikolaj@91.225.203.16) Quit (Quit: away)
[23:28] * dscastro (~dscastro@190.18.48.177) Quit (Remote host closed the connection)
[23:33] * efirs (~firs@c-50-185-70-125.hsd1.ca.comcast.net) Quit (Ping timeout: 480 seconds)
[23:33] * mhack (~mhack@66-168-117-78.dhcp.oxfr.ma.charter.com) Quit (Remote host closed the connection)
[23:42] * sleinen (~Adium@2001:620:1000:3:a65e:60ff:fedb:f305) Quit (Quit: Leaving.)
[23:48] * dscastro (~dscastro@190.18.48.177) has joined #ceph
[23:49] * JohnO (~Jourei@149.91.83.212) has joined #ceph
[23:52] * Discovery (~Discovery@178.239.49.70) Quit (Read error: Connection reset by peer)
[23:59] * InIMoeK (~InIMoeK@029-226-128-083.dynamic.caiway.nl) has joined #ceph

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