#ceph IRC Log

Index

IRC Log for 2011-02-24

Timestamps are in GMT/BST.

[0:00] <cmccabe> rather weird default
[0:00] <cmccabe> what else about it is BS though
[0:00] <gregaf> it defaults to mounting the root Ceph dir...
[0:01] <gregaf> I don't remember what the 'r' is for, actually
[0:01] <cmccabe> oh, I see. So it's the root of the Ceph FS, not a client path
[0:01] <gregaf> yeah
[0:01] <cmccabe> the r means that anyone using common_init will have the -r option set g_conf.client_mountpoint
[0:02] <gregaf> oh right
[0:02] <Tv> cmccabe: it makes e.g. cfuse start without mountpoint arg
[0:02] <cmccabe> which might be what tv is objecting to
[0:02] <Tv> it should not have a default value in the first place
[0:02] <gregaf> Tv: why on earth wouldn't you want it to?
[0:02] <Tv> oh that's inside the ceph fs
[0:02] <Tv> ok then i got mistracked chasing the argv's
[0:02] <gregaf> it doesn't start without a directory to mount to :)
[0:02] <cmccabe> heh
[0:03] <Tv> gregaf: yes it does
[0:03] <cmccabe> chasing the argvs...
[0:03] <gregaf> well it errors out
[0:03] <gregaf> gregf@kai:~/ceph/src$ ./cfuse
[0:03] <gregaf> 2011-02-23 15:18:58.716772 7ff01826b720 ceph version 0.25~rc (commit:e4b8274241df201e01f2280d9f9e3857b4265f87)
[0:03] <gregaf> cfuse[15202]: starting ceph client
[0:03] <gregaf> cfuse[15202]: starting fuse
[0:03] <gregaf> fuse: missing mountpoint parameter
[0:03] <gregaf> cfuse[15202]: fuse finished with error -2
[0:03] <gregaf> cfuse[15201]: mount failed: No such file or directory
[0:04] <Tv> cfuse -c /dev/null -m 127.0.0.1:1234
[0:04] <gregaf> does it still return 0 or something?
[0:04] <gregaf> anyway, gotta run
[0:04] <Tv> gregaf: if it errs out for you, that means it talks to monitor before parsing command line args
[0:04] * Yoric (~David@dau94-10-88-189-211-192.fbx.proxad.net) Quit (Quit: Yoric)
[0:04] <Tv> which is bs
[0:04] <cmccabe> tv: cfuse parses its arguments inside common_init like all the other daemons
[0:05] <cmccabe> tv: well... it parses some of them that way; the rest it does in that loop you can see in main
[0:05] <Tv> cmccabe: ehh, it doesn't take the exact same options as everything else
[0:05] <cmccabe> tv: in either case, you won't be sending any traffic until you see the line // start up network
[0:05] <cmccabe> SimpleMessenger *messenger = new SimpleMessenger();
[0:05] <Tv> anyway, that command, with nothing on port 1234, will hang, not err out
[0:06] <cmccabe> tv: I think the only flag cfuse takes that is different from everyone else is --localize-reads
[0:06] <Tv> i have no idea where it'd try to mount things if there was a valid ceph cluster to talk to
[0:07] <Tv> (or if it'd err out at that point)
[0:07] <Tv> (or crash)
[0:07] <cmccabe> tv: oh, actually there is something wacky about cfuse
[0:07] <cmccabe> tv: it takes a third swipe at argv inside the fuse stuff
[0:08] <cmccabe> tv: inside ceph_fuse_ll_main
[0:10] <cmccabe> tv: I think that's mostly so users can pass options straight through into fuse. FUSE has some options of its own that aren't ceph-related
[0:10] <cmccabe> tv: like -o allow_other, -o max_read=N, etc.
[0:11] <Tv> yeah all i can say is i'm trying to put together automated tests so i can rip that code a new one and feel safe
[0:11] <cmccabe> tv: haha
[0:11] <cmccabe> tv: passing the extra args through to fuse is the right approach, though. As clumsy as it may seem.
[0:12] <Tv> yeah, but failing to notice a missing mandatory arg is not cool
[0:12] <cmccabe> tv: FUSE has a bunch of random options that sometimes need to be tweaked
[0:12] * Meths_ is now known as Meths
[0:12] <cmccabe> tv: but since we're already cooking the argv beforehand, perhaps we should do a better job at it
[0:12] <cmccabe> tv: and at least notice the lack of mountpoint
[0:18] * greglap (~Adium@166.205.138.87) has joined #ceph
[0:32] <cmccabe> does anyone object if I merge the librados_api_config branch back into librados_api? Then I'd like to finish with the librados python/C++ bindings
[0:37] <joshd> cmccabe: sounds good to me
[0:37] * alexxy (~alexxy@79.173.81.171) Quit (Ping timeout: 480 seconds)
[0:45] * allsystemsarego (~allsystem@188.25.128.78) Quit (Quit: Leaving)
[1:14] * greglap (~Adium@166.205.138.87) Quit (Read error: Connection reset by peer)
[1:59] <Tv> how do i negate a single-character option with this command line arg parser? it's an OPT_BOOL that defaults to true and i want to set it to false
[2:01] <cmccabe> tv: which parser is "this parser"? The config.cc one?
[2:02] <Tv> in this case, cfuse -d
[2:03] <cmccabe> yeah, that is ceph_argparse
[2:04] <cmccabe> well, you could pass cfuse -f
[2:04] <Tv> } else if (CONF_ARG_EQ("nodaemon", 'D')) {
[2:04] <Tv> hmms
[2:04] <cmccabe> -D is like -f, but with more spew
[2:04] <cmccabe> as in stderr spew
[2:04] <Tv> naturally -f is not documented anywhere
[2:05] <cmccabe> generic_usage
[2:05] <cmccabe> all programs that use ceph_argparse should be showing generic_usage as part of their usage
[2:05] <cmccabe> of course in practice... they don't.
[2:05] <Tv> snafu
[2:06] <cmccabe> I actually don't know how OPT_BOOL options with a char set are parsed on the command-line; I use them so rarely
[2:07] <cmccabe> actually there's only one that has ever existed: -d
[2:07] <cmccabe> OPTION(daemonize, 'd', OPT_BOOL, false),
[2:08] <cmccabe> does -d take an argument? This is a mystery I do not know
[2:08] <Tv> -d0 did not work
[2:08] <Tv> anyway, i'll happily loop back to fixing all that once we have a comfy pile of tests
[2:09] <cmccabe> it definitely wouldn't be -d0. If anything, it would be -d false
[2:09] <Tv> yeah i tried that because something def'ed some BOOL to INT
[2:09] <cmccabe> mm.. I thought they were mostly separate... bool handles true and false at least
[2:09] <cmccabe> anyway, I have to finish some other stuff, so I'll have to figure that out another day
[2:21] * Tv (~Tv|work@ip-66-33-206-8.dreamhost.com) Quit (Ping timeout: 480 seconds)
[2:32] * ooolinux (~bless@211.144.206.177) has joined #ceph
[2:33] * bless_ (~bless@211.144.206.177) has joined #ceph
[2:34] * bless_ (~bless@211.144.206.177) Quit ()
[2:34] <ooolinux> ?
[2:39] * ooolinux (~bless@211.144.206.177) Quit (Quit: Leaving)
[3:30] * joshd (~joshd@ip-66-33-206-8.dreamhost.com) Quit (Quit: Leaving.)
[3:37] * cmccabe (~cmccabe@c-24-23-253-6.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[3:50] * greglap (~Adium@cpe-76-90-239-202.socal.res.rr.com) has joined #ceph
[4:14] * ooolinux (~bless@211.144.206.177) has joined #ceph
[4:15] * ooolinux (~bless@211.144.206.177) Quit ()
[4:35] * johnl is now known as Guest2449
[5:42] * eternaleye (~eternaley@195.215.30.181) Quit (Ping timeout: 480 seconds)
[5:46] * morse (~morse@supercomputing.univpm.it) Quit (Ping timeout: 480 seconds)
[5:47] * eternaleye (~eternaley@195.215.30.181) has joined #ceph
[6:33] * greglap (~Adium@cpe-76-90-239-202.socal.res.rr.com) Quit (Quit: Leaving.)
[6:33] * greglap (~Adium@cpe-76-90-239-202.socal.res.rr.com) has joined #ceph
[7:58] * Yoric (~David@dau94-10-88-189-211-192.fbx.proxad.net) has joined #ceph
[8:13] * bchrisman (~Adium@c-98-207-207-62.hsd1.ca.comcast.net) has joined #ceph
[9:07] * allsystemsarego (~allsystem@188.25.128.78) has joined #ceph
[9:09] * Yoric (~David@dau94-10-88-189-211-192.fbx.proxad.net) Quit (Quit: Yoric)
[9:53] * verwilst (~verwilst@router.begen1.office.netnoc.eu) has joined #ceph
[10:01] * Yoric (~David@213.144.210.93) has joined #ceph
[10:12] * alexxy (~alexxy@79.173.81.171) has joined #ceph
[10:52] * morse (~morse@supercomputing.univpm.it) has joined #ceph
[14:09] * alexxy (~alexxy@79.173.81.171) Quit (Remote host closed the connection)
[14:09] * Psi-Jack_ (~psi-jack@yggdrasil.hostdruids.com) Quit (Quit: Leaving)
[14:16] * alexxy (~alexxy@79.173.81.171) has joined #ceph
[14:17] * ghaskins_mobile (~ghaskins_@66-189-113-47.dhcp.oxfr.ma.charter.com) has joined #ceph
[14:41] * ghaskins_mobile (~ghaskins_@66-189-113-47.dhcp.oxfr.ma.charter.com) Quit (Quit: This computer has gone to sleep)
[17:45] * Yoric (~David@213.144.210.93) Quit (Quit: Yoric)
[17:56] * bchrisman (~Adium@c-98-207-207-62.hsd1.ca.comcast.net) Quit (Quit: Leaving.)
[17:56] * verwilst (~verwilst@router.begen1.office.netnoc.eu) Quit (Quit: Ex-Chat)
[18:00] * joshd (~joshd@ip-66-33-206-8.dreamhost.com) has joined #ceph
[18:18] * Tv (~Tv|work@ip-66-33-206-8.dreamhost.com) has joined #ceph
[18:35] * alexxy (~alexxy@79.173.81.171) Quit (Ping timeout: 480 seconds)
[18:36] * cmccabe (~cmccabe@c-24-23-253-6.hsd1.ca.comcast.net) has joined #ceph
[18:38] * alexxy (~alexxy@79.173.81.171) has joined #ceph
[18:38] * bchrisman (~Adium@70-35-37-146.static.wiline.com) has joined #ceph
[18:39] * ana_ (~ana@mini.inf.um.es) has joined #ceph
[18:42] * Yoric (~David@dau94-10-88-189-211-192.fbx.proxad.net) has joined #ceph
[18:45] * alexxy (~alexxy@79.173.81.171) Quit (Quit: No Ping reply in 180 seconds.)
[18:48] * alexxy (~alexxy@79.173.81.171) has joined #ceph
[19:18] <wido> sagewk: Do you have a moment?
[19:19] <wido> In a about 2 ~ 3 weeks I'll have my new cluster up and running and I'm preparing the test I'm going to run. Are there some specific things which need testing?
[19:30] <Tv> sagewk: gregaf: yehudasa: sjust: 10:30?
[19:31] <sjust> Tv: sure
[19:32] <cmccabe> k
[19:38] <Tv> it seems sagewk has disappeared
[19:38] <Tv> i guess we'll postpone..
[19:50] <sagewk> i'm back!
[20:41] * Guest2449 (~johnl@johnl.ipq.co) Quit (Remote host closed the connection)
[20:41] * johnl (~johnl@johnl.ipq.co) has joined #ceph
[20:42] * WesleyS (~WesleyS@12.248.40.138) has joined #ceph
[20:44] * Yoric (~David@dau94-10-88-189-211-192.fbx.proxad.net) Quit (Ping timeout: 480 seconds)
[20:46] * cclien (~cclien@ec2-175-41-146-71.ap-southeast-1.compute.amazonaws.com) Quit (Remote host closed the connection)
[20:46] * cclien (~cclien@ec2-175-41-146-71.ap-southeast-1.compute.amazonaws.com) has joined #ceph
[20:49] * WesleyS (~WesleyS@12.248.40.138) has left #ceph
[20:59] * Juul (~Juul@slim.dhcp.lbl.gov) has joined #ceph
[20:59] * Juul (~Juul@slim.dhcp.lbl.gov) Quit (Remote host closed the connection)
[20:59] * Juul (~Juul@space.labitat.dk) has joined #ceph
[21:02] * Juul (~Juul@space.labitat.dk) Quit (Remote host closed the connection)
[21:02] * Juul (~Juul@static.88-198-13-205.clients.your-server.de) has joined #ceph
[21:36] * verwilst (~verwilst@dD576FAAE.access.telenet.be) has joined #ceph
[22:08] <sjust> su
[22:08] <sjust> oops
[22:08] <cmccabe> localhost #
[22:08] <Tv> make me a sandwich
[22:09] <sjust> since when does becoming super user cause the machine to start issuing *you* commands?
[22:12] <Tv> i was expecting something more like "make: *** No rule to make target `me'. Stop."
[22:25] <Tv> 13:25:31 DEBUG| Running '/usr/local/autotest/tests/bonnie/src/bonnie++ -d /usr/local/autotest/tmp/tmpDKT7or_cfuse_iozone.cluster0/mnt.0 -u root -n 2048'
[22:25] <Tv> mwahahaha
[22:26] <Tv> repurposing existing autotest fs tests to run on ceph
[22:44] * greglap (~Adium@cpe-76-90-239-202.socal.res.rr.com) Quit (Ping timeout: 480 seconds)
[22:45] * prometheanfire (~mthode@mx1.mthode.org) Quit (Ping timeout: 480 seconds)
[22:45] * Meths (rift@91.106.223.232) Quit (Ping timeout: 480 seconds)
[22:45] * greglap (~Adium@cpe-76-90-239-202.socal.res.rr.com) has joined #ceph
[22:48] * prometheanfire (~mthode@mx1.mthode.org) has joined #ceph
[22:51] * Meths (rift@91.106.223.232) has joined #ceph
[23:08] * bchrisman (~Adium@70-35-37-146.static.wiline.com) Quit (Quit: Leaving.)
[23:48] * verwilst (~verwilst@dD576FAAE.access.telenet.be) Quit (Quit: Ex-Chat)
[23:55] <Tv> hrmmph.. i have an almost idle ceph cluster, and a very slow bonnie++ benchmark
[23:56] <Tv> i wonder if the virtualization crap is causing a huge disk write overhead
[23:56] <sagewk> which part of the benchmark is it?
[23:56] <Tv> 13:25:31 ERROR| [stderr] Using uid:0, gid:0.
[23:56] <Tv> 13:42:36 ERROR| [stderr] Writing with putc()...done
[23:56] <Tv> 13:59:33 ERROR| [stderr] Writing intelligently...done
[23:56] <Tv> 14:15:21 ERROR| [stderr] Rewriting...done
[23:56] <Tv> 14:17:27 ERROR| [stderr] Reading with getc()...done
[23:56] <Tv> 14:18:06 ERROR| [stderr] Reading intelligently...done
[23:56] <Tv> 14:19:52 ERROR| [stderr] start 'em...done...done...done...
[23:56] <Tv> (yeah, look at 'em timestamps!)
[23:56] <Tv> this is in http://autotest.ceph.newdream.net/results/161-tv/group0/testvm-008.ceph.newdream.net/debug/client.0.log
[23:58] <darkfader> Tv: i hope you don't got qemu dm somewhere in that?
[23:59] <Tv> darkfader: kvm and virtio disks
[23:59] <darkfader> (ah lol. right. you needed a reasonably current kernel anyway)
[23:59] * ghaskins_mobile (~ghaskins_@66-189-113-47.dhcp.oxfr.ma.charter.com) has joined #ceph

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