/srv/irclogs.ubuntu.com/2014/07/03/#cloud-init.txt

=== harlowja is now known as harlowja_away
=== evilissimo|afk is now known as evilissimo
=== zz_gondoi is now known as gondoi
=== evilissimo is now known as evilissimo|afk
=== evilissimo|afk is now known as evilissimo
=== alexpilotti_ is now known as alexpilotti
mnaserIt looks like cloud-init backgrounded resize is blocking.. and any ideas why it can be taking a really long time? Jul  3 14:43:13 localhost [CLOUDINIT] util.py[DEBUG]: backgrounded Resizing took 275.751 seconds14:47
mnaserThat's from 20G to 160GB (with only ~6GB of data)14:47
mnaserThis is on CentOS 6, upgraded to the latest e2fsprogs14:48
harmwthat kernel doesn't support it, iirc14:48
mnaserharmw: afaik, there is a feature that makes resizing much faster using lazy_itable_init14:49
harmwcentos' kernel is to old to certain resize things14:49
mnaserhm.14:49
harmwi haven't looked at this for say 4 months or so though :)14:49
mnaserAny suggestions in that case? I rather not drop in a kernel that won't get updated, any "updated centos kernel" repos?14:49
harmwno no, kernels are holy14:50
mnaseryeah.. i didn't for a while but we had someone provision a 1.2tb disk centos vm and it's taking forever and a half :)14:50
harmwperhaps centos-plus, though I doubt even that14:50
harmwwell, why on earth would one require a cloudy instance with 1.2TB of storage?14:50
harmw:)14:50
harmw(and ofcourse, 'because one can' is a viable answer :p)14:51
mnaserharmw: our local disk grows with the amount of memry/etc14:51
mnaserso this is a 32gb instance14:51
mnaserex: 1gb gets 40gb... we grow linearly so 32 ends up with 1280gb14:51
harmwdamn14:52
mnaseryeah but this isnt too useful if i cant figure out how we can solve this14:52
mnaseri tried e2fsprogs 1.42.6, 1.42.8 and 1.42.10 (latest) with the utils as well but no go14:52
harmwon a personal note, since when did there become a direct relation between mem/hdd sizes? :)14:52
harmwthats not a very good thing imho14:52
mnaserhow come?  it helps us not end up with un-used local disk space14:53
harmwanyway, problem here is resizing takes quite long due to the large disk/fs14:53
mnaseryeah.. i think it has to do with the old kernel/tooling14:53
harmwI'm quite positive it was the old (ancient) CentOS kernel in this case14:53
harmwhave you tried ubuntu?14:54
harmwor fedora?14:54
mnaserubuntu boots ultra quick with no problems14:54
harmwand thats with kernel 3.x?14:54
mnaserthat's why i felt it had to be resize2fs/kernel14:54
mnaseryup14:54
harmwyeah, in that case I'm quite positive this boiled down to CentOS' kernel14:54
mnaseri found this too14:54
mnaserhttps://bugs.launchpad.net/bugs/117961014:54
harmwshouldnt take to look for them to set 7.0 free though :)14:55
mnaserit looks like ubuntu had the same issue in 201314:55
harmwyea, we had this issue with cirros a while back as well14:55
mnaserit does look like this got fixed by newer e2fsprogs but that didnt help me14:56
harmwnah, it still requires a shiny 3.x kernel14:57
* mnaser sigh14:57
harmwyou might want to give centos7 a try, it's not far from gold status14:58
mnaserpretty sure the stuff we run on centos 6 hasnt yet been fully prepped for centos 714:59
mnaserhttp://elrepo.org/tiki/kernel-lt i want to give this a shot14:59
harmwah yes15:01
harmwwell good luck15:02
mnaserharmw: thanks, entering uncharted territory is fun (not)15:02
harmwhehe15:02
mnaseri'll let you know though! :)15:02
harmwyea, I'd appreciate that15:02
=== evilissimo is now known as evilissimo|afk
=== praneshp_ is now known as praneshp
=== gondoi is now known as zz_gondoi
=== zz_gondoi is now known as gondoi
=== gondoi is now known as zz_gondoi
=== zz_gondoi is now known as gondoi
=== gondoi is now known as zz_gondoi

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!