Re: CEPH OSD Restarts taking too long v10.2.9

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



The issue we have is large leveldb's . do we have any setting to disable compaction of leveldb on osd start?


On Fri, Mar 29, 2019 at 7:44 PM Nikhil R <nikh.ravindra@xxxxxxxxx> wrote:
Any help on this would be much appreciated as our prod is down since a day and each osd restart is taking 4-5 hours.


On Fri, Mar 29, 2019 at 7:43 PM Nikhil R <nikh.ravindra@xxxxxxxxx> wrote:
We have maxed out the files per dir. CEPH is trying to do an online split due to which osd's are crashing. We increased the split_multiple and merge_threshold for now and are restarting osd's. Now on these restarts the leveldb compaction is taking a long time. Below are some of the logs. 

2019-03-29 06:25:37.082055 7f3c6320a8c0  0 genericfilestorebackend(/var/lib/ceph/osd/ceph-83) detect_features: FIEMAP ioctl is disabled via 'filestore fiemap' config option
2019-03-29 06:25:37.082064 7f3c6320a8c0  0 genericfilestorebackend(/var/lib/ceph/osd/ceph-83) detect_features: SEEK_DATA/SEEK_HOLE is disabled via 'filestore seek data hole' config option
2019-03-29 06:25:37.082079 7f3c6320a8c0  0 genericfilestorebackend(/var/lib/ceph/osd/ceph-83) detect_features: splice is supported
2019-03-29 06:25:37.096658 7f3c6320a8c0  0 genericfilestorebackend(/var/lib/ceph/osd/ceph-83) detect_features: syncfs(2) syscall fully supported (by glibc and kernel)
2019-03-29 06:25:37.096703 7f3c6320a8c0  0 xfsfilestorebackend(/var/lib/ceph/osd/ceph-83) detect_feature: extsize is disabled by conf
2019-03-29 06:25:37.295577 7f3c6320a8c0  1 leveldb: Recovering log #1151738
2019-03-29 06:25:37.445516 7f3c6320a8c0  1 leveldb: Delete type=0 #1151738
2019-03-29 06:25:37.445574 7f3c6320a8c0  1 leveldb: Delete type=3 #1151737
2019-03-29 07:11:50.619313 7ff6c792b700  1 leveldb: Compacting 1@3 + 12@4 files
2019-03-29 07:11:50.639795 7ff6c792b700  1 leveldb: Generated table #1029200: 7805 keys, 2141956 bytes
2019-03-29 07:11:50.649315 7ff6c792b700  1 leveldb: Generated table #1029201: 4464 keys, 1220994 bytes
2019-03-29 07:11:50.660485 7ff6c792b700  1 leveldb: Generated table #1029202: 7813 keys, 2142882 bytes
2019-03-29 07:11:50.672235 7ff6c792b700  1 leveldb: Generated table #1029203: 6283 keys, 1712810 bytes
2019-03-29 07:11:50.697949 7ff6c792b700  1 leveldb: Generated table #1029204: 7805 keys, 2142841 bytes
2019-03-29 07:11:50.714648 7ff6c792b700  1 leveldb: Generated table #1029205: 5173 keys, 1428905 bytes
2019-03-29 07:11:50.757146 7ff6c792b700  1 leveldb: Generated table #1029206: 7888 keys, 2143304 bytes
2019-03-29 07:11:50.774357 7ff6c792b700  1 leveldb: Generated table #1029207: 5168 keys, 1425634 bytes
2019-03-29 07:11:50.830276 7ff6c792b700  1 leveldb: Generated table #1029208: 7821 keys, 2146114 bytes
2019-03-29 07:11:50.849116 7ff6c792b700  1 leveldb: Generated table #1029209: 6106 keys, 1680947 bytes
2019-03-29 07:11:50.909866 7ff6c792b700  1 leveldb: Generated table #1029210: 7799 keys, 2142782 bytes
2019-03-29 07:11:50.921143 7ff6c792b700  1 leveldb: Generated table #1029211: 5737 keys, 1574963 bytes
2019-03-29 07:11:50.923357 7ff6c792b700  1 leveldb: Generated table #1029212: 1149 keys, 310202 bytes
2019-03-29 07:11:50.923388 7ff6c792b700  1 leveldb: Compacted 1@3 + 12@4 files => 22214334 bytes
2019-03-29 07:11:50.924224 7ff6c792b700  1 leveldb: compacted to: files[ 0 3 54 715 6304 24079 0 ]
2019-03-29 07:11:50.942586 7ff6c792b700  1 leveldb: Delete type=2 #1029109

Is there a way i can skip this?



On Fri, Mar 29, 2019 at 11:32 AM huang jun <hjwsm1989@xxxxxxxxx> wrote:
Nikhil R <nikh.ravindra@xxxxxxxxx> 于2019年3月29日周五 下午1:44写道:
>
> if i comment filestore_split_multiple = 72 filestore_merge_threshold = 480   in the ceph.conf wont ceph take the default value of 2 and 10 and we would be in more splits and crashes?
>
Yes, that aimed to make it clear what results in the long start time,
leveldb compact or filestore split?
> in.linkedin.com/in/nikhilravindra
>
>
>
> On Fri, Mar 29, 2019 at 6:55 AM huang jun <hjwsm1989@xxxxxxxxx> wrote:
>>
>> It seems like the split settings result the problem,
>> what about comment out those settings then see it still used that long
>> time to restart?
>> As a fast search in code, these two
>> filestore_split_multiple = 72
>> filestore_merge_threshold = 480
>> doesn't support online change.
>>
>> Nikhil R <nikh.ravindra@xxxxxxxxx> 于2019年3月28日周四 下午6:33写道:
>> >
>> > Thanks huang for the reply.
>> > Its is the disk compaction taking more time
>> > the disk i/o is completely utilized upto 100%
>> > looks like both osd_compact_leveldb_on_mount = false & leveldb_compact_on_mount = false isnt working as expected on ceph v10.2.9
>> > is there a way to turn off compaction?
>> >
>> > Also, the reason why we are restarting osd's is due to splitting and we increased split multiple and merge_threshold.
>> > Is there a way we would inject it? Is osd restarts the only solution?
>> >
>> > Thanks In Advance
>> >
>> > in.linkedin.com/in/nikhilravindra
>> >
>> >
>> >
>> > On Thu, Mar 28, 2019 at 3:58 PM huang jun <hjwsm1989@xxxxxxxxx> wrote:
>> >>
>> >> Did the time really cost on db compact operation?
>> >> or you can turn on debug_osd=20 to see what happens,
>> >> what about the disk util during start?
>> >>
>> >> Nikhil R <nikh.ravindra@xxxxxxxxx> 于2019年3月28日周四 下午4:36写道:
>> >> >
>> >> > CEPH osd restarts are taking too long a time
>> >> > below is my ceph.conf
>> >> > [osd]
>> >> > osd_compact_leveldb_on_mount = false
>> >> > leveldb_compact_on_mount = false
>> >> > leveldb_cache_size=1073741824
>> >> > leveldb_compression = false
>> >> > osd_mount_options_xfs = "rw,noatime,inode64,logbsize=256k"
>> >> > osd_max_backfills = 1
>> >> > osd_recovery_max_active = 1
>> >> > osd_recovery_op_priority = 1
>> >> > filestore_split_multiple = 72
>> >> > filestore_merge_threshold = 480
>> >> > osd_max_scrubs = 1
>> >> > osd_scrub_begin_hour = 22
>> >> > osd_scrub_end_hour = 3
>> >> > osd_deep_scrub_interval = 2419200
>> >> > osd_scrub_sleep = 0.1
>> >> >
>> >> > looks like both osd_compact_leveldb_on_mount = false & leveldb_compact_on_mount = false isnt working as expected on ceph v10.2.9
>> >> >
>> >> > Any ideas on a fix would be appreciated asap
>> >> > in.linkedin.com/in/nikhilravindra
>> >> >
>> >> > _______________________________________________
>> >> > ceph-users mailing list
>> >> > ceph-users@xxxxxxxxxxxxxx
>> >> > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>> >>
>> >>
>> >>
>> >> --
>> >> Thank you!
>> >> HuangJun
>>
>>
>>
>> --
>> Thank you!
>> HuangJun



--
Thank you!
HuangJun
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Ceph Dev]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]


  Powered by Linux