#ceph IRC Log

Index

IRC Log for 2011-06-24

Timestamps are in GMT/BST.

[0:00] <Tv> i always explain missing public transport with lag
[0:00] <Tv> or walking into walls
[0:03] <darkfader> Tv: idspispopd
[0:04] <stingray> idchoppers
[0:06] * allsystemsarego (~allsystem@188.27.164.204) Quit (Quit: Leaving)
[0:28] * aliguori (~anthony@32.97.110.65) has joined #ceph
[1:09] <stingray> ugh
[1:15] <stingray> everything is worse than I expected.
[1:15] * stingray going to sleep
[1:23] * aliguori (~anthony@32.97.110.65) Quit (Ping timeout: 480 seconds)
[2:02] * Tv (~Tv|work@ip-66-33-206-8.dreamhost.com) Quit (Ping timeout: 480 seconds)
[2:12] * sugoruyo (~george@athedsl-408632.home.otenet.gr) Quit (Quit: sugoruyo)
[2:15] * greglap (~Adium@ip-66-33-206-8.dreamhost.com) Quit (Quit: Leaving.)
[2:24] * yoshi (~yoshi@p24092-ipngn1301marunouchi.tokyo.ocn.ne.jp) has joined #ceph
[2:38] * cmccabe (~cmccabe@c-24-23-254-199.hsd1.ca.comcast.net) has left #ceph
[2:48] * joshd (~joshd@ip-66-33-206-8.dreamhost.com) Quit (Quit: Leaving.)
[3:12] * bchrisman (~Adium@70-35-37-146.static.wiline.com) Quit (Quit: Leaving.)
[3:24] * jim (~chatzilla@astound-69-42-16-6.ca.astound.net) Quit (Remote host closed the connection)
[3:27] * jim (~chatzilla@astound-69-42-16-6.ca.astound.net) has joined #ceph
[5:05] * Dantman (~dantman@S0106001731dfdb56.vs.shawcable.net) has joined #ceph
[5:35] * greglap (~Adium@cpe-76-170-84-245.socal.res.rr.com) has joined #ceph
[5:39] * bchrisman (~Adium@c-98-207-207-62.hsd1.ca.comcast.net) has joined #ceph
[6:47] * greglap (~Adium@cpe-76-170-84-245.socal.res.rr.com) Quit (Quit: Leaving.)
[7:00] * greglap (~Adium@cpe-76-170-84-245.socal.res.rr.com) has joined #ceph
[7:13] * Yulya__ (~Yu1ya_@ip-95-220-158-194.bb.netbynet.ru) has joined #ceph
[7:15] * Yulya_ (~Yu1ya_@ip-95-220-161-118.bb.netbynet.ru) Quit (Ping timeout: 480 seconds)
[8:00] * Dantman (~dantman@S0106001731dfdb56.vs.shawcable.net) Quit (Ping timeout: 480 seconds)
[8:04] * Dantman (~dantman@S0106001731dfdb56.vs.shawcable.net) has joined #ceph
[9:12] * verwilst_ (~verwilst@dD576903F.access.telenet.be) has joined #ceph
[11:08] * sugoruyo (~george@athedsl-408632.home.otenet.gr) has joined #ceph
[11:45] * yoshi (~yoshi@p24092-ipngn1301marunouchi.tokyo.ocn.ne.jp) Quit (Remote host closed the connection)
[11:59] <stingray> so I reset the journal and now I have much more data than I have metadata for
[11:59] <stingray> there's no way I can clear orphan objects?
[12:14] * stingray will try stupid - find . -printf %i\\n. then convert to hex and delete all that's not in list from data
[12:20] <sugoruyo> is there some way i can find out the objects a file has been striped to?
[12:24] <stingray> get the inode number, convert to hex, then rados -p data ls | grep ${CONVERTEDTOHEX}
[12:24] <stingray> at least this will be my guess
[12:25] <stingray> to verify, export them, cat them together, and see what you get
[12:32] * Yulya__ (~Yu1ya_@ip-95-220-158-194.bb.netbynet.ru) Quit (Ping timeout: 480 seconds)
[14:51] * aliguori (~anthony@cpe-70-123-132-139.austin.res.rr.com) has joined #ceph
[14:56] * aliguori (~anthony@cpe-70-123-132-139.austin.res.rr.com) Quit (Quit: Ex-Chat)
[14:59] * mtk (~mtk@ool-182c8e6c.dyn.optonline.net) Quit (Remote host closed the connection)
[15:05] * mtk (~mtk@ool-182c8e6c.dyn.optonline.net) has joined #ceph
[15:11] * lx0 (~aoliva@186.214.53.147) has joined #ceph
[15:18] * lxo (~aoliva@83TAAB0X0.tor-irc.dnsbl.oftc.net) Quit (Ping timeout: 480 seconds)
[15:23] * sugoruyo (~george@athedsl-408632.home.otenet.gr) Quit (Read error: Connection reset by peer)
[15:23] * sugoruyo (~george@athedsl-408632.home.otenet.gr) has joined #ceph
[16:03] * aliguori (~anthony@32.97.110.65) has joined #ceph
[16:11] * morse (~morse@supercomputing.univpm.it) Quit (Remote host closed the connection)
[16:57] * yoshi (~yoshi@KD027091032046.ppp-bb.dion.ne.jp) has joined #ceph
[16:57] * yoshi (~yoshi@KD027091032046.ppp-bb.dion.ne.jp) Quit (Remote host closed the connection)
[17:02] * morse (~morse@supercomputing.univpm.it) has joined #ceph
[17:30] * verwilst_ (~verwilst@dD576903F.access.telenet.be) Quit (Quit: Ex-Chat)
[17:34] * greglap (~Adium@cpe-76-170-84-245.socal.res.rr.com) Quit (Quit: Leaving.)
[17:47] * bchrisman (~Adium@c-98-207-207-62.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[17:59] * greglap (~Adium@166.205.137.242) has joined #ceph
[18:17] * greglap (~Adium@166.205.137.242) Quit (Quit: Leaving.)
[18:38] * bchrisman (~Adium@70-35-37-146.static.wiline.com) has joined #ceph
[18:38] * Tv (~Tv|work@ip-66-33-206-8.dreamhost.com) has joined #ceph
[18:58] * aliguori (~anthony@32.97.110.65) Quit (Quit: Ex-Chat)
[19:02] * cmccabe (~cmccabe@208.80.64.174) has joined #ceph
[19:04] * joshd (~joshd@ip-66-33-206-8.dreamhost.com) has joined #ceph
[19:07] * greglap (~Adium@ip-66-33-206-8.dreamhost.com) has joined #ceph
[19:12] * aliguori (~anthony@32.97.110.65) has joined #ceph
[19:54] * rsharpe (~Adium@70-35-37-146.static.wiline.com) Quit (Ping timeout: 480 seconds)
[19:54] * bchrisman (~Adium@70-35-37-146.static.wiline.com) Quit (Ping timeout: 480 seconds)
[19:55] * rsharpe (~Adium@70-35-37-146.static.wiline.com) has joined #ceph
[20:09] * bchrisman (~Adium@70-35-37-146.static.wiline.com) has joined #ceph
[20:21] * sagelap (~sage@ip-66-33-206-8.dreamhost.com) has joined #ceph
[20:27] <cmccabe> just fyi, I am bisecting the testrados issue now
[20:37] * lx0 (~aoliva@186.214.53.147) Quit (Read error: Connection reset by peer)
[20:38] * lx0 (~aoliva@9YYAABOUR.tor-irc.dnsbl.oftc.net) has joined #ceph
[20:47] <sjust> cmccabe: 965aadc4c7d33d2332fdf6b3995880b10ffb8cd2 is a likely candidate
[20:47] <cmccabe> sjust: I have 3 steps left to test
[20:47] <sjust> ok
[20:48] <cmccabe> what does watch/notify have to do with snaps by the way?
[20:48] <sjust> absolutely nothing
[20:48] <cmccabe> that is something I never understood
[20:48] <sjust> well, as little as any two concepts in the osd have to do with each other
[20:49] <cmccabe> heh
[20:49] <sjust> I should have committed that on it's own, I am very bad :(
[20:49] <cmccabe> so you can wait for changes to an object then
[20:49] <cmccabe> or watch the object
[20:49] <cmccabe> and get a notifier callback
[20:49] <sjust> thats the idea
[20:49] <sjust> I don't know much about watch/notify
[20:50] <sjust> I just know that I modified some code around it
[20:50] <sjust> it *should* have been ok
[20:50] <cmccabe> testceph is also failing, but I think that might be the same problem
[20:50] <cmccabe> so I'm not going to debug it until we at least know what's causing this problem
[20:50] <sjust> yup
[20:50] <sjust> I was just volunteering information
[20:51] <cmccabe> yeah
[20:57] <cmccabe> sigh
[20:57] <cmccabe> it bisected to 629ac0da7adeb314c19d1e22deccc11534dbd0e6
[20:57] <cmccabe> which does... nothing. Except remove one call to derr()
[20:58] * lx0 (~aoliva@9YYAABOUR.tor-irc.dnsbl.oftc.net) Quit (Read error: No route to host)
[20:58] <yehudasa> cmccabe:
[20:58] <yehudasa> cmccabe: that's a bug actually...
[20:59] <stingray> what's so special in inode 100 ?
[20:59] <cmccabe> I'm going to try again with make distclean before every build
[20:59] <joshd> this is why we should put braces around 1-line ifs
[20:59] <yehudasa> cmccabe: before you do that, remove the if that preceded that derr
[20:59] <cmccabe> yehudasa: whether or not you liked the derr(), I don't think you can argue that removing it broken watch/notify
[20:59] <cmccabe> oh... wait!
[21:00] <cmccabe> yeah, there it is
[21:01] <cmccabe> man, I really hate 2-space indentation
[21:01] <cmccabe> thanks for spotting that yehuda
[21:01] <yehudasa> np
[21:03] * lxo (~aoliva@83TAAB1X7.tor-irc.dnsbl.oftc.net) has joined #ceph
[21:23] <gregaf1> stingray: inode 100 is the first of the special mds dirs
[21:24] <stingray> okay
[21:24] <cmccabe> now bisecting testceph
[21:24] <stingray> is it possible to explain what's so special in it in a couple of sentences, or I shall read the code?
[21:25] <stingray> I was about to try walking through the regular directory inodes
[21:25] <stingray> that is, for the real lost file finder
[21:26] <stingray> I kinda did what I want, roughly, by getting all inodes and then wiping all extra stuff from data pool
[21:26] <stingray> but that's too lame
[21:26] <cmccabe> stingray: this is about the time greg normally goes to lunch
[21:27] <gregaf1> stingray: so each logical MDS has a data directory that contains its journal and stray inodes (roughly, those inodes that still have caps out but are no longer linked into the MDS tree)
[21:27] <gregaf1> that data directory is ino 100 for mds 0, 101 for mds1, etc
[21:27] <cmccabe> gregaf: oh, didn't know if you were there.
[21:27] <gregaf1> this is when I'm usually back from lunch, although sometimes it can be longer :)
[21:28] <gregaf1> I don't recall offhand what the journal inos are, but I think they might start at 200?
[21:28] * mtk (~mtk@ool-182c8e6c.dyn.optonline.net) Quit (Remote host closed the connection)
[21:28] <stingray> gregaf1: aha. okay.
[21:28] <stingray> there are also 6xx something inodes
[21:29] * mtk (5Y0khGggq6@panix2.panix.com) has joined #ceph
[21:29] <stingray> anyway I guess I can ignore those for now and just start with object "1"
[21:29] <stingray> I was also looking for a way to tell mds to flush its journal
[21:30] <stingray> and by flush I mean to play it onto the inodes to the end
[21:30] <stingray> so then I can kill it and ignore the journal but still get the correct info
[21:30] <stingray> I couldn't find it :(
[21:30] <gregaf1> stingray: I don't think there's a good way to do that right now
[21:31] <gregaf1> after all, if the journal can be replayed then there's no need to kill it
[21:31] * sagelap (~sage@ip-66-33-206-8.dreamhost.com) Quit (Quit: Leaving.)
[21:32] <stingray> yeah, except for the cases when it cannot
[21:32] <stingray> which I happen to encounter a lot
[21:33] * lxo (~aoliva@83TAAB1X7.tor-irc.dnsbl.oftc.net) Quit (Read error: No route to host)
[21:33] <gregaf1> yeah, that's because you're running with a bazillion MDSes, that code's still getting stabilized
[21:33] <stingray> anyway, don't bother. if I'll get to it I'll just try adding a command for tell
[21:33] <gregaf1> but still, that means it can't replay the journal, so a flush command wouldn't do much good
[21:33] <stingray> hehe :)
[21:34] <stingray> the last case it was 1 mds
[21:34] <stingray> 1 mon
[21:34] <stingray> and 2 osds
[21:34] <gregaf1> last case?
[21:34] <gregaf1> that one from yesterday that you gave me the journal for?
[21:34] <stingray> yep.
[21:34] <gregaf1> ungh
[21:34] <gregaf1> well, that just moved up in priority
[21:35] <stingray> :|
[21:36] * lxo (~aoliva@1GLAACGHM.tor-irc.dnsbl.oftc.net) has joined #ceph
[21:46] * greglap (~Adium@ip-66-33-206-8.dreamhost.com) Quit (Quit: Leaving.)
[21:54] * mtk (5Y0khGggq6@panix2.panix.com) Quit (Remote host closed the connection)
[21:55] * mtk (~mtk@ool-182c8e6c.dyn.optonline.net) has joined #ceph
[21:55] * mtk (~mtk@ool-182c8e6c.dyn.optonline.net) Quit ()
[21:56] * mtk (~mtk@ool-182c8e6c.dyn.optonline.net) has joined #ceph
[22:01] * macana (~ml.macana@159.226.41.129) Quit ()
[22:03] * lxo (~aoliva@1GLAACGHM.tor-irc.dnsbl.oftc.net) Quit (Ping timeout: 480 seconds)
[22:17] * lxo (~aoliva@09GAAE4QJ.tor-irc.dnsbl.oftc.net) has joined #ceph
[22:21] <Tv> you just gotta love a project where the last year of development is 15 commits with the following unique lines of commit messages:
[22:21] <Tv> .
[22:21] <Tv> support for xml/jtl output
[22:21] <Tv> re-process results with -r command line option
[22:39] <stingray> on my projects my commit messages are almost exclusively "."
[22:39] <stingray> with the occasional exception of "WIP", "It works, yay!" and "WTF"
[22:49] * jantje (~jan@paranoid.nl) has joined #ceph
[22:56] * sugoruyo (~george@athedsl-408632.home.otenet.gr) Quit (charon.oftc.net kinetic.oftc.net)
[22:56] * wido_ (~wido@fubar.widodh.nl) Quit (charon.oftc.net kinetic.oftc.net)
[22:56] * jantje_ (~jan@paranoid.nl) Quit (charon.oftc.net kinetic.oftc.net)
[22:56] * wonko_be (bernard@november.openminds.be) Quit (charon.oftc.net kinetic.oftc.net)
[22:56] * Anticimex (anticimex@netforce.csbnet.se) Quit (charon.oftc.net kinetic.oftc.net)
[22:56] * wonko_be (bernard@november.openminds.be) has joined #ceph
[22:58] * sugoruyo (~george@athedsl-408632.home.otenet.gr) has joined #ceph
[22:58] * wido_ (~wido@fubar.widodh.nl) has joined #ceph
[22:58] * Anticimex (anticimex@netforce.csbnet.se) has joined #ceph
[23:09] * sagelap (~sage@ip-66-33-206-8.dreamhost.com) has joined #ceph
[23:10] <failbaitr> what styleguide do you guys use anyway?
[23:11] <sjust> heh
[23:11] <sjust> you mean coding style?
[23:11] <sjust> we are talking about adopting a version of google's C++ guide
[23:15] * lxo (~aoliva@09GAAE4QJ.tor-irc.dnsbl.oftc.net) Quit (Read error: Connection reset by peer)
[23:20] <failbaitr> sjust: yeah
[23:20] <failbaitr> Im guessing you use some normalized style so everyone adheres to the same rules?
[23:20] <sjust> somewhat
[23:20] <sjust> but it's not formalized yet
[23:20] <failbaitr> Ah ok
[23:21] <sjust> it's on our todo list :)
[23:21] <failbaitr> Ill see when it happend
[23:21] <cmccabe> I think we actually put it on to the list of things that will get done in the next few weeks
[23:21] <sjust> ah, good
[23:22] <cmccabe> but I forget exactly where it was on that list
[23:28] * lxo (~aoliva@9YYAABOY5.tor-irc.dnsbl.oftc.net) has joined #ceph
[23:37] * aliguori (~anthony@32.97.110.65) Quit (Read error: Operation timed out)
[23:52] * aliguori (~anthony@32.97.110.65) has joined #ceph
[23:59] * bchrisman (~Adium@70-35-37-146.static.wiline.com) Quit (Quit: Leaving.)

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