#ceph IRC Log

Index

IRC Log for 2010-08-16

Timestamps are in GMT/BST.

[0:06] * ghaskins_mobile (~ghaskins_@66-189-114-103.dhcp.oxfr.ma.charter.com) Quit (Quit: This computer has gone to sleep)
[1:12] * ghaskins_mobile (~ghaskins_@66-189-114-103.dhcp.oxfr.ma.charter.com) has joined #ceph
[1:44] * ghaskins_mobile (~ghaskins_@66-189-114-103.dhcp.oxfr.ma.charter.com) Quit (Quit: This computer has gone to sleep)
[1:46] * ghaskins_mobile (~ghaskins_@66-189-114-103.dhcp.oxfr.ma.charter.com) has joined #ceph
[3:14] * ghaskins_mobile (~ghaskins_@66-189-114-103.dhcp.oxfr.ma.charter.com) Quit (Quit: This computer has gone to sleep)
[4:30] * ghaskins_mobile (~ghaskins_@66-189-114-103.dhcp.oxfr.ma.charter.com) has joined #ceph
[6:40] * jbl (~jbl@charybdis-ext.suse.de) Quit (Ping timeout: 480 seconds)
[6:56] * f4m8_ is now known as f4m8
[7:29] * jbl (~jbl@charybdis-ext.suse.de) has joined #ceph
[8:17] * allsystemsarego (~allsystem@188.26.33.211) has joined #ceph
[8:19] * Osso (osso@AMontsouris-755-1-2-32.w86-212.abo.wanadoo.fr) Quit (Quit: Osso)
[11:46] * lidongyang (~lidongyan@222.126.194.154) Quit (Remote host closed the connection)
[11:49] * lidongyang (~lidongyan@222.126.194.154) has joined #ceph
[13:16] * MarkN (~nathan@59.167.240.178) Quit (Ping timeout: 480 seconds)
[13:32] * MarkN (~nathan@59.167.240.178) has joined #ceph
[15:34] * josef (~seven@nat-pool-rdu.redhat.com) has joined #ceph
[15:34] <josef> sage: i remember getting an email from you but i'll be damned if i can find it
[15:37] <josef> sage: oh ok i found it, I can make it so debuginfo isn't stripped for the entirety of ceph, does that work for you?
[15:41] * josef waits for the sun to catch up to the west coast
[15:57] * f4m8 is now known as f4m8_
[16:46] * lore (~ask@pool-96-228-112-217.albyny.east.verizon.net) has joined #ceph
[16:50] * lore (~ask@pool-96-228-112-217.albyny.east.verizon.net) has left #ceph
[16:50] * gregphone (~gregphone@166.205.137.46) has joined #ceph
[16:50] * lore (~ask@pool-96-228-112-217.albyny.east.verizon.net) has joined #ceph
[16:52] <gregphone> josef: pretty sure that's good, we like useful stack traces from all our code ;)
[16:57] <sagewk> josef: yeah, skipping the strip for the whole package is fine :)
[17:17] <todinini> I tried the qemu rbd branch, but I get this error
[17:17] <todinini> root@node43:~# ./qemu-img create -f rbd rbd:data/foo1 10G
[17:17] <todinini> Formatting 'rbd:data/foo1', fmt=rbd size=10737418240 cluster_size=0
[17:17] <todinini> failed assigning block id
[17:17] <todinini> qemu-img: rbd:data/foo1: error while creating rbd: Input/output error
[17:18] <todinini> sagewk: the stalling write on /sys/class/rbd/add from friday is fixed. thanks
[17:20] <sagewk> todinini: what does 'ceph class list' say?
[17:21] <sagewk> you may need to 'cclass -a' to load the latest rbd class into the cluster.. the block id stuff is somewhat newish..?
[17:23] <todinini> root@node43:~# ceph class list
[17:23] <todinini> 10.08.16_17:22:53.898202 mon <- [class,list]
[17:23] <todinini> 10.08.16_17:22:53.899163 mon1 -> 'installed classes:
[17:23] <todinini> rbd (v1.2 [x86-64]) [active]
[17:23] <todinini> ' (0)
[17:24] <todinini> ohh, now it works, cclass -a needs a moment
[17:25] <sagewk> cool
[17:31] <josef> sagewk: ok i will do that then :)
[17:31] <sagewk> thanks
[17:32] <sagewk> let me know if you see any issues with the .spec, a lot of random people submitted patches (and i don't build rpms myself)
[17:34] <josef> sagewk: will do
[17:51] * gregphone (~gregphone@166.205.137.46) Quit (Ping timeout: 480 seconds)
[18:42] <wido> hi
[18:42] <sagewk> wido: hi
[18:42] <wido> i think i found a possible file corruption: http://www.pastebin.org/557542
[18:43] <wido> using the unstable branch on the clients
[18:43] <sagewk> yep, ok. probably related to the rbd osdc changes. are those files still there on your cluster?
[18:45] <wido> which files? i just ran this test, a few minutes ago
[18:45] <sagewk> the good and bad copies of that iso
[18:45] <wido> oh, yes, they are still there
[18:53] * Osso (osso@AMontsouris-755-1-31-136.w90-46.abo.wanadoo.fr) has joined #ceph
[19:00] <wido> anyone here using make-kpkg for building the ceph-client? make-kpkg keeps adding a + as suffix to my kernel package name: linux-image-2.6.35-rbd+
[19:02] <sagewk> iirc that's a new thing with 2.6.35 (or .34).. the + tellsyou it's not vanilla 2.6.35.
[19:03] <sagewk> you can turn it off with some argument to make (VER_SUFFIX or something like that)
[19:04] <wido> ok, but make-kpkg keeps breaking on that, since the package names do not match
[19:04] <wido> 2.6.35-rbd+ is not in control info
[19:06] <sagewk> wido: ok, osd7 disk is full, and the error isn't propagating.
[19:06] <wido> oh, i didn't notice that indeed?
[19:07] <wido> it's a 94%, just 1% below the threshold
[19:12] <gregaf> yeah, and btrfs has similar issues of its own with space so it's returning ENOSPC on some operations, before the drive is actually full
[19:19] <wido> yes, i read something about that
[19:21] <sagewk> defragging hte btrfs btree.. that should help
[19:23] <wido> but sagewk the corruption is due to a filled up osd?
[19:24] <sagewk> yeah, the object write fails, but cosd isn't noticing, and when it goes to read it back it's 0 bytes.
[19:25] <sagewk> so cosd needs to do something less stupid, and btrfs should report ENOSPC less often :)
[19:29] <wido> ok, shall i create an issue for it?
[19:35] <sagewk> yes please
[19:36] <sagewk> actually, #354 should cover it
[19:53] <wido> ah, indeed. Tnx gregaf ;)
[20:49] <lore> does ceph need a logo?
[20:49] <sagewk> i suppose that's a matter of opinion
[20:50] <lore> i mean if you're interested let me know
[20:50] <gregaf> by which he means, does http://ceph.newdream.net/images/ceph-logo1.jpg count
[20:50] <lore> ive been working on perl's rakudo
[20:50] <lore> just doing logos for opensource apps/etc
[20:52] <sagewk> as in, http://rakudo.org/sites/default/files/logo.png ?
[20:54] <lore> nay
[20:54] <lore> i personally dislike that
[20:55] <lore> im modifying their current one for rakudostar
[20:55] <lore> something along the lines of http://i38.tinypic.com/14l3i1h.jpg
[20:55] <lore> still modifying the color on it
[20:56] <gregaf> uh, there is no color in that image?
[20:56] <lore> there is
[20:56] <lore> but i think they want morecolor to the tree
[22:29] * lore (~ask@pool-96-228-112-217.albyny.east.verizon.net) Quit ()
[22:43] * allsystemsarego (~allsystem@188.26.33.211) Quit (Quit: Leaving)

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