Re: kraken-bluestore 11.2.0 memory leak issue

nokia ceph <nokiacephusers@xxxxxxxxx> · Wed, 29 Mar 2017 12:42:31 +0530

Hello,
We manually fixed the issue and below is  our analysis.

Due to high CPU utilisation we stopped the ceph-mgr on all our cluster.
On one of our cluster we saw high memory usage by OSDs some grater than 5GB causing OOM , resulting in process kill.
The memory was released immediately when the ceph-mgr started . So, this issue is clearly a side effect of stopping the ceph-mgr process . What we dont understand is why all OSDs are reporting a single OSD issue and locking so much of memory until ceph-mgr is started ?
1. Ceph status higliting there is issue on one of the OSD in 5th node "wrong node":
cn1.vn1ldv1c1.cdn ~# ceph status
2017-03-28 05:54:52.210450 7f8108a84700 1 WARNING: the following dangerous and experimental features are enabled: bluestore,rocksdb
2017-03-28 05:54:52.231551 7f8108a84700 -1 WARNING: the following dangerous and experimental features are enabled: bluestore,rocksdb
2017-03-28 05:54:52.400565 7f8101ac6700 0 - 10.139.4.81:0/2856869581 >> 10.139.4.85:6800/44958 conn(0x7f80e8002bc0 :-1 s=STATE_CONNECTING_WAIT_BANNER_AND_IDENTIFY pgs=0 cs=0 l=1)._process_connection connect claims to be 10.139.4.85:6800/273761 not 10.139.4.85:6800/44958 - wrong node!
cluster 71a32568-96f3-4998-89dd-e2e7d77a6824
health HEALTH_OK
monmap e3: 5 mons at {cn1=10.139.4.81:6789/0,cn2=10.139.4.82:6789/0,cn3=10.139.4.83:6789/0,cn4=10.139.4.84:6789/0,cn5=10.139.4.85:6789/0}
election epoch 24, quorum 0,1,2,3,4 cn1,cn2,cn3,cn4,cn5
mgr active: cn5
osdmap e2010: 335 osds: 335 up, 335 in
flags sortbitwise,require_jewel_osds,require_kraken_osds
pgmap v561117: 8192 pgs, 1 pools, 28323 GB data, 12400 kobjects
37667 GB used, 1182 TB / 1218 TB avail
8192 active+clean
client io 31732 kB/s rd, 57763 kB/s wr, 59 op/s rd, 479 op/s wr
2. numastat of ceph shows that it has consumed total of 275GB memory and most of them consuming more than 5GB.
cn1.vn1ldv1c1.cdn /var/log/cassandra# numastat -s ceph
Per-node process memory usage (in MBs)
PID Node 0 Node 1 Total
----------------- --------------- --------------- ---------------
372602 (ceph-osd) 5418.34 2.84 5421.18
491602 (ceph-osd) 5351.95 2.83 5354.78
417717 (ceph-osd) 5175.98 2.83 5178.81
273980 (ceph-osd) 5167.83 2.82 5170.65
311956 (ceph-osd) 5167.04 2.84 5169.88
440537 (ceph-osd) 5161.57 2.84 5164.41
368422 (ceph-osd) 5157.87 2.83 5160.70
292227 (ceph-osd) 5156.42 2.83 5159.25
360749 (ceph-osd) 5129.43 2.83 5132.26
516040 (ceph-osd) 5112.53 2.84 5115.37
526274 (ceph-osd) 5108.76 2.83 5111.59
300197 (ceph-osd) 5096.64 2.82 5099.46
487087 (ceph-osd) 5081.70 2.82 5084.52
396562 (ceph-osd) 5060.55 2.84 5063.38
409201 (ceph-osd) 5058.06 2.83 5060.89
284767 (ceph-osd) 5027.94 2.82 5030.76
520653 (ceph-osd) 4997.16 2.82 4999.98
302873 (ceph-osd) 4988.78 2.82 4991.60
364601 (ceph-osd) 4884.61 2.83 4887.43
426998 (ceph-osd) 4865.89 2.82 4868.72
294856 (ceph-osd) 4851.96 2.82 4854.78
306064 (ceph-osd) 4780.84 2.85 4783.68
449676 (ceph-osd) 4764.82 2.84 4767.66
376469 (ceph-osd) 4750.47 2.82 4753.29
482502 (ceph-osd) 4729.85 2.84 4732.69
357126 (ceph-osd) 4706.88 2.82 4709.71
346001 (ceph-osd) 4693.43 2.84 4696.27
511640 (ceph-osd) 4668.58 2.82 4671.41
282682 (ceph-osd) 4614.66 2.84 4617.50
287330 (ceph-osd) 4613.75 2.82 4616.57
506197 (ceph-osd) 4604.59 2.84 4607.43
332253 (ceph-osd) 4587.28 2.82 4590.11
413603 (ceph-osd) 4579.29 2.84 4582.12
297473 (ceph-osd) 4569.20 2.84 4572.04
431396 (ceph-osd) 4537.83 2.84 4540.66
501424 (ceph-osd) 4533.19 2.84 4536.03
477729 (ceph-osd) 4505.37 2.83 4508.20
392520 (ceph-osd) 4439.75 2.84 4442.59
280349 (ceph-osd) 4389.06 2.82 4391.88
321805 (ceph-osd) 4385.80 2.82 4388.62
463759 (ceph-osd) 4369.09 2.82 4371.91
328971 (ceph-osd) 4349.35 2.82 4352.18
530916 (ceph-osd) 4330.21 2.82 4333.03
468626 (ceph-osd) 4272.68 2.83 4275.51
353261 (ceph-osd) 4266.01 2.82 4268.83
339729 (ceph-osd) 4194.93 2.82 4197.75
422844 (ceph-osd) 4157.31 2.82 4160.14
400631 (ceph-osd) 4155.34 2.82 4158.16
325467 (ceph-osd) 4144.66 2.84 4147.50
380309 (ceph-osd) 4119.42 2.82 4122.24
454764 (ceph-osd) 4007.09 2.82 4009.92
336089 (ceph-osd) 4003.25 2.82 4006.07
349613 (ceph-osd) 3953.32 2.84 3956.15
473107 (ceph-osd) 3833.75 2.83 3836.59
388421 (ceph-osd) 3776.79 2.83 3779.62
308957 (ceph-osd) 3758.94 2.82 3761.76
315430 (ceph-osd) 3677.42 2.82 3680.24
445064 (ceph-osd) 3669.27 2.82 3672.09
977162 (ceph-osd) 1508.02 3.40 1511.43
166155 (ceph-osd) 1411.64 3.42 1415.06
228123 (ceph-osd) 1399.20 3.41 1402.60
39367 (ceph-osd) 1397.44 3.41 1400.85
228124 (ceph-osd) 1227.50 3.41 1230.91
284384 (ceph-osd) 1204.96 3.41 1208.37
339890 (ceph-osd) 1139.69 3.41 1143.10
467652 (ceph-osd) 1016.18 3.41 1019.59
597584 (ceph-osd) 901.18 3.41 904.58
934986 (ceph-mon) 0.02 184.65 184.67
----------------- --------------- --------------- ---------------
Total 278720.27 379.42 279099.69
3. OOM tries to kill the ceph-osd process
Mar 27 01:57:05 cn1 kernel: ceph-osd invoked oom-killer: gfp_mask=0x280da, order=0, oom_score_adj=0
Mar 27 01:57:05 cn1 kernel: ceph-osd cpuset=/ mems_allowed=0-1
Mar 27 01:57:05 cn1 kernel: CPU: 0 PID: 422861 Comm: ceph-osd Not tainted 3.10.0-327.el7.x86_64 #1
Mar 27 01:57:05 cn1 kernel: Hardware name: HP ProLiant XL450 Gen9 Server/ProLiant XL450 Gen9 Server, BIOS U21 09/12/2016
Mar 27 01:57:05 cn1 kernel: ffff884546751700 00000000275e2e50 ffff88454137b6f0 ffffffff816351f1
Mar 27 01:57:05 cn1 kernel: ffff88454137b780 ffffffff81630191 000000000487ffff ffff8846f0665590
Mar 27 01:57:05 cn1 kernel: ffff8845411b3ad8 ffff88454137b7d0 ffffffffffffffd5 0000000000000001
Mar 27 01:57:05 cn1 kernel: Call Trace:
Mar 27 01:57:05 cn1 kernel: [<ffffffff816351f1>] dump_stack+0x19/0x1b
Mar 27 01:57:05 cn1 kernel: [<ffffffff81630191>] dump_header+0x8e/0x214
Mar 27 01:57:05 cn1 kernel: [<ffffffff8116cdee>] oom_kill_process+0x24e/0x3b0
Mar 27 01:57:05 cn1 kernel: [<ffffffff8116c956>] ? find_lock_task_mm+0x56/0xc0
Mar 27 01:57:05 cn1 kernel: [<ffffffff8116d616>] out_of_memory+0x4b6/0x4f0
Mar 27 01:57:05 cn1 kernel: [<ffffffff811737f5>] __alloc_pages_nodemask+0xa95/0xb90
Mar 27 01:57:05 cn1 kernel: [<ffffffff811b78ca>] alloc_pages_vma+0x9a/0x140
Mar 27 01:57:05 cn1 kernel: [<ffffffff81197655>] handle_mm_fault+0xb85/0xf50
Mar 27 01:57:05 cn1 kernel: [<ffffffffa04f5b22>] ? xfs_perag_get_tag+0x42/0xe0 [xfs]
Mar 27 01:57:05 cn1 kernel: [<ffffffff81640e22>] __do_page_fault+0x152/0x420
Mar 27 01:57:05 cn1 kernel: [<ffffffff81641113>] do_page_fault+0x23/0x80
Mar 27 01:57:05 cn1 kernel: [<ffffffff8163d408>] page_fault+0x28/0x30
Mar 27 01:57:05 cn1 kernel: [<ffffffff813000c9>] ? copy_user_enhanced_fast_string+0x9/0x20
Mar 27 01:57:05 cn1 kernel: [<ffffffff8130600a>] ? memcpy_toiovec+0x4a/0x90
Mar 27 01:57:05 cn1 kernel: [<ffffffff8151f91f>] skb_copy_datagram_iovec+0x12f/0x2a0
Mar 27 01:57:05 cn1 kernel: [<ffffffff81574418>] tcp_recvmsg+0x248/0xbc0
Mar 27 01:57:05 cn1 kernel: [<ffffffff810bb685>] ? sched_clock_cpu+0x85/0xc0
Mar 27 01:57:05 cn1 kernel: [<ffffffff815a10eb>] inet_recvmsg+0x7b/0xa0
Mar 27 01:57:05 cn1 kernel: [<ffffffff8150ffb6>] sock_aio_read.part.7+0x146/0x160
Mar 27 01:57:05 cn1 kernel: [<ffffffff8150fff1>] sock_aio_read+0x21/0x30
Mar 27 01:57:05 cn1 kernel: [<ffffffff811ddcdd>] do_sync_read+0x8d/0xd0
Mar 27 01:57:05 cn1 kernel: [<ffffffff811de4e5>] vfs_read+0x145/0x170
Mar 27 01:57:05 cn1 kernel: [<ffffffff811def8f>] SyS_read+0x7f/0xe0
Mar 27 01:57:05 cn1 kernel: [<ffffffff81645909>] system_call_fastpath+0x16/0x1b
Mar 27 01:57:05 cn1 kernel: Mem-Info:
Mar 27 01:57:05 cn1 kernel: Node 0 DMA per-cpu:
4. On all OSDs below error flooded
2017-03-28 12:51:28.889658 7f82dd053700 0 -- 10.139.4.83:6850/121122 >> 10.139.4.85:6800/44958 conn(0x7f82eeeea000 :-1 s=STATE_CONNECTING_WAIT_BANNER_AND_IDENTIFY pgs=1164657 cs=2 l=0)._process_connection connect claims to be 10.139.4.85:6800/273761 not 10.139.4.85:6800/44958 - wrong node!
on affected osd logs,
2017-03-28 12:51:29.191346 7f8a775b6700 0 -- 10.139.4.85:6800/273761 >> - conn(0x7f8aad6c0000 :6800 s=STATE_ACCEPTING_WAIT_BANNER_ADDR pgs=0 cs=0 l=0).fault with nothing to send and in the half accept state just closed
2017-03-28 12:51:29.249841 7f8a775b6700 0 -- 10.139.4.85:6800/273761 >> - conn(0x7f8aacc34800 :6800 s=STATE_ACCEPTING_WAIT_BANNER_ADDR pgs=0 cs=0 l=0).fault with nothing to send and in the half accept state just closed
5. Once ceph-mgr started on the affected OSD node, all OSDs move to reconnect state.
2017-03-28 12:52:37.468455 7f8a77db7700 0 -- 10.139.4.85:6800/273761 >> 10.139.4.85:6928/185705 conn(0x7f8aacc34800 :-1 s=STATE_OPEN pgs=32 cs=1 l=0).*fault initiating reconnect*
2017-03-28 12:52:37.468502 7f2bbcd5b700 0 -- 10.139.4.85:6897/154091 >> 10.139.4.85:6928/185705 conn(0x7f2c63448800 :-1 s=STATE_OPEN pgs=301 cs=1 l=0).fault initiating reconnect
2017-03-28 12:52:37.469503 7fd36f161700 0 -- 10.139.4.84:6822/95096 >> 10.139.4.85:6928/185705 conn(0x7fd412f69800 :-1 s=STATE_OPEN pgs=173 cs=1 l=0).*fault initiating reconnect*
2017-03-28 12:52:37.463913 7f82dd053700 0 -- 10.139.4.83:6850/121122 >> 10.139.4.85:6928/185705 conn(0x7f83da64b800 :-1 s=STATE_OPEN pgs=154 cs=1 l=0).*fault initiating reconnect*
2017-03-28 12:52:37.468406 7fea1fc05700 0 -- 10.139.4.82:6816/97331 >> 10.139.4.85:6928/185705 conn(0x7feab70f6000 :-1 s=STATE_OPEN pgs=108 cs=1 l=0).*fault initiating reconnect*
6. Suddenly the memory decreased from 275GB to 147GB.
So, what is the impact ceph-mgr creating here ?

Thanks

On Tue, Mar 28, 2017 at 2:49 PM, Jay Linux <jaylinuxgeek@xxxxxxxxx> wrote:
Hello,
We still facing same memory leak issue even if we specify bluestore_cache_size to 100M which caused ceph OSD process killed by out of memory . 

Mar 27 01:57:05 cn1 kernel: ceph-osd invoked oom-killer: gfp_mask=0x280da, order=0, oom_score_adj=0
Mar 27 01:57:05 cn1 kernel: ceph-osd cpuset=/ mems_allowed=0-1
Mar 27 01:57:05 cn1 kernel: CPU: 0 PID: 422861 Comm: ceph-osd Not tainted 3.10.0-327.el7.x86_64 #1
Mar 27 01:57:05 cn1 kernel: Hardware name: HP ProLiant XL450 Gen9 Server/ProLiant XL450 Gen9 Server, BIOS U21 09/12/2016
Mar 27 01:57:05 cn1 kernel: ffff884546751700 00000000275e2e50 ffff88454137b6f0 ffffffff816351f1
Mar 27 01:57:05 cn1 kernel: ffff88454137b780 ffffffff81630191 000000000487ffff ffff8846f0665590
Mar 27 01:57:05 cn1 kernel: ffff8845411b3ad8 ffff88454137b7d0 ffffffffffffffd5 0000000000000001
Mar 27 01:57:05 cn1 kernel: Call Trace:
Mar 27 01:57:05 cn1 kernel: [<ffffffff816351f1>] dump_stack+0x19/0x1b
Mar 27 01:57:05 cn1 kernel: [<ffffffff81630191>] dump_header+0x8e/0x214
Mar 27 01:57:05 cn1 kernel: [<ffffffff8116cdee>] oom_kill_process+0x24e/0x3b0
Mar 27 01:57:05 cn1 kernel: [<ffffffff8116c956>] ? find_lock_task_mm+0x56/0xc0
Mar 27 01:57:05 cn1 kernel: [<ffffffff8116d616>] out_of_memory+0x4b6/0x4f0
Mar 27 01:57:05 cn1 kernel: [<ffffffff811737f5>] __alloc_pages_nodemask+0xa95/0xb90
Mar 27 01:57:05 cn1 kernel: [<ffffffff811b78ca>] alloc_pages_vma+0x9a/0x140
Mar 27 01:57:05 cn1 kernel: [<ffffffff81197655>] handle_mm_fault+0xb85/0xf50
Mar 27 01:57:05 cn1 kernel: [<ffffffffa04f5b22>] ? xfs_perag_get_tag+0x42/0xe0 [xfs]
Mar 27 01:57:05 cn1 kernel: [<ffffffff81640e22>] __do_page_fault+0x152/0x420
Mar 27 01:57:05 cn1 kernel: [<ffffffff81641113>] do_page_fault+0x23/0x80
Mar 27 01:57:05 cn1 kernel: [<ffffffff8163d408>] page_fault+0x28/0x30
Mar 27 01:57:05 cn1 kernel: [<ffffffff813000c9>] ? copy_user_enhanced_fast_string+0x9/0x20
Mar 27 01:57:05 cn1 kernel: [<ffffffff8130600a>] ? memcpy_toiovec+0x4a/0x90
Mar 27 01:57:05 cn1 kernel: [<ffffffff8151f91f>] skb_copy_datagram_iovec+0x12f/0x2a0
Mar 27 01:57:05 cn1 kernel: [<ffffffff81574418>] tcp_recvmsg+0x248/0xbc0
Mar 27 01:57:05 cn1 kernel: [<ffffffff810bb685>] ? sched_clock_cpu+0x85/0xc0
Mar 27 01:57:05 cn1 kernel: [<ffffffff815a10eb>] inet_recvmsg+0x7b/0xa0
Mar 27 01:57:05 cn1 kernel: [<ffffffff8150ffb6>] sock_aio_read.part.7+0x146/0x160
Mar 27 01:57:05 cn1 kernel: [<ffffffff8150fff1>] sock_aio_read+0x21/0x30
Mar 27 01:57:05 cn1 kernel: [<ffffffff811ddcdd>] do_sync_read+0x8d/0xd0
Mar 27 01:57:05 cn1 kernel: [<ffffffff811de4e5>] vfs_read+0x145/0x170
Mar 27 01:57:05 cn1 kernel: [<ffffffff811def8f>] SyS_read+0x7f/0xe0
Mar 27 01:57:05 cn1 kernel: [<ffffffff81645909>] system_call_fastpath+0x16/0x1b

For several occurance of OOM event

#dmesg -T | grep -i memory
[Mon Mar 27 02:51:25 2017] Out of memory: Kill process 459076 (ceph-osd) score 18 or sacrifice child
[Mon Mar 27 06:41:16 2017]  [<ffffffff8116d616>] out_of_memory+0x4b6/0x4f0
[Mon Mar 27 06:41:16 2017] Out of memory: Kill process 976901 (java) score 31 or sacrifice child
[Mon Mar 27 06:43:55 2017]  [<ffffffff8116d616>] out_of_memory+0x4b6/0x4f0
[Mon Mar 27 06:43:55 2017] Out of memory: Kill process 37351 (java) score 31 or sacrifice child
[Mon Mar 27 06:43:55 2017]  [<ffffffff8116d616>] out_of_memory+0x4b6/0x4f0
[Mon Mar 27 06:43:55 2017] Out of memory: Kill process 435981 (ceph-osd) score 17 or sacrifice child
[Mon Mar 27 11:06:07 2017]  [<ffffffff8116d616>] out_of_memory+0x4b6/0x4f0

# numactl -H
available: 2 nodes (0-1)
node 0 cpus: 0 1 2 3 4 5 6 7 16 17 18 19 20 21 22 23
node 0 size: 294786 MB
node 0 free: 3447 MB  ===>>> Used almost 98%

While analysing numastat results, here you can find each osd consumes more than 5G. 

====
# numastat -s ceph

Per-node process memory usage (in MBs)
PID                         Node 0          Node 1           Total
-----------------  --------------- --------------- ---------------
372602 (ceph-osd)          5418.34            2.84         5421.18  
491602 (ceph-osd)          5351.95            2.83         5354.78
417717 (ceph-osd)          5175.98            2.83         5178.81
273980 (ceph-osd)          5167.83            2.82         5170.65
311956 (ceph-osd)          5167.04            2.84         5169.88
440537 (ceph-osd)          5161.57            2.84         5164.41
368422 (ceph-osd)          5157.87            2.83         5160.70
292227 (ceph-osd)          5156.42            2.83         5159.25
====

Is there any way to fix the memory leak? Awaiting for your comments. 

---
bluestore_cache_size = 107374182
bluefs_buffered_io = true
---

Env:- RHEL7.2
         v11.2.0 kraken , EC 4+1

FYI - http://tracker.ceph.com/issues/18924 Raised already a tracker for this issue.

Thanks

On Mon, Feb 20, 2017 at 11:18 AM, Jay Linux <jaylinuxgeek@xxxxxxxxx> wrote:
Hello Shinobu,
We already raised ticket for this issue. FYI - http://tracker.ceph.com/issues/18924

Thanks
Jayaram

On Mon, Feb 20, 2017 at 12:36 AM, Shinobu Kinjo <skinjo@xxxxxxxxxx> wrote:
Please open ticket at http://tracker.ceph.com, if you haven't yet.

On Thu, Feb 16, 2017 at 6:07 PM, Muthusamy Muthiah

<muthiah.muthusamy@xxxxxxxxx> wrote:

> Hi Wido,

>

> Thanks for the information and let us know if this is a bug.

> As workaround we will go with small bluestore_cache_size to 100MB.

>

> Thanks,

> Muthu

>

> On 16 February 2017 at 14:04, Wido den Hollander <wido@xxxxxxxx> wrote:

>>

>>

>> > Op 16 februari 2017 om 7:19 schreef Muthusamy Muthiah

>> > <muthiah.muthusamy@xxxxxxxxx>:

>> >

>> >

>> > Thanks IIya Letkowski for the information we will change this value

>> > accordingly.

>> >

>>

>> What I understand from yesterday's performance meeting is that this seems

>> like a bug. Lowering this buffer reduces memory, but the root-cause seems to

>> be memory not being freed. A few bytes of a larger allocation still

>> allocated causing this buffer not to be freed.

>>

>> Tried:

>>

>> debug_mempools = true

>>

>> $ ceph daemon osd.X dump_mempools

>>

>> Might want to view the YouTube video of yesterday when it's online:

>> https://www.youtube.com/channel/UCno-Fry25FJ7B4RycCxOtfw/videos

>>

>> Wido

>>

>> > Thanks,

>> > Muthu

>> >

>> > On 15 February 2017 at 17:03, Ilya Letkowski <mj12.svetzari@xxxxxxxxx>

>> > wrote:

>> >

>> > > Hi, Muthusamy Muthiah

>> > >

>> > > I'm not totally sure that this is a memory leak.

>> > > We had same problems with bluestore on ceph v11.2.0.

>> > > Reduce bluestore cache helped us to solve it and stabilize OSD memory

>> > > consumption on the 3GB level.

>> > >

>> > > Perhaps this will help you:

>> > >

>> > > bluestore_cache_size = 104857600

>> > >

>> > >

>> > >

>> > > On Tue, Feb 14, 2017 at 11:52 AM, Muthusamy Muthiah <

>> > > muthiah.muthusamy@xxxxxxxxx> wrote:

>> > >

>> > >> Hi All,

>> > >>

>> > >> On all our 5 node cluster with ceph 11.2.0 we encounter memory leak

>> > >> issues.

>> > >>

>> > >> Cluster details : 5 node with 24/68 disk per node , EC : 4+1 , RHEL

>> > >> 7.2

>> > >>

>> > >> Some traces using sar are below and attached the memory utilisation

>> > >> graph

>> > >> .

>> > >>

>> > >> (16:54:42)[cn2.c1 sa] # sar -r

>> > >> 07:50:01 kbmemfree kbmemused %memused kbbuffers kbcached kbcommit

>> > >> %commit

>> > >> kbactive kbinact kbdirty

>> > >> 10:20:01 32077264 132754368 80.54 16176 3040244 77767024 47.18

>> > >> 51991692

>> > >> 2676468 260

>> > >>

>> > >>

>> > >>

>> > >>

>> > >>

>> > >>

>> > >>

>> > >>

>> > >> *10:30:01 32208384 132623248 80.46 16176 3048536 77832312 47.22

>> > >> 51851512

>> > >> 2684552 1210:40:01 32067244 132764388 80.55 16176 3059076 77832316

>> > >> 47.22

>> > >> 51983332 2694708 26410:50:01 30626144 134205488 81.42 16176 3064340

>> > >> 78177232 47.43 53414144 2693712 411:00:01 28927656 135903976 82.45

>> > >> 16176

>> > >> 3074064 78958568 47.90 55114284 2702892 1211:10:01 27158548 137673084

>> > >> 83.52

>> > >> 16176 3080600 80553936 48.87 56873664 2708904 1211:20:01 26455556

>> > >> 138376076

>> > >> 83.95 16176 3080436 81991036 49.74 57570280 2708500 811:30:01

>> > >> 26002252

>> > >> 138829380 84.22 16176 3090556 82223840 49.88 58015048 2718036

>> > >> 1611:40:01

>> > >> 25965924 138865708 84.25 16176 3089708 83734584 50.80 58049980

>> > >> 2716740

>> > >> 1211:50:01 26142888 138688744 84.14 16176 3089544 83800100 50.84

>> > >> 57869628

>> > >> 2715400 16*

>> > >>

>> > >> ...

>> > >> ...

>> > >>

>> > >> In the attached graph, there is increase in memory utilisation by

>> > >> ceph-osd during soak test. And when it reaches the system limit of

>> > >> 128GB

>> > >> RAM , we could able to see the below dmesg logs related to memory out

>> > >> when

>> > >> the system reaches close to 128GB RAM. OSD.3 killed due to Out of

>> > >> memory

>> > >> and started again.

>> > >>

>> > >> [Tue Feb 14 03:51:02 2017] *tp_osd_tp invoked oom-killer:

>> > >> gfp_mask=0x280da, order=0, oom_score_adj=0*

>> > >> [Tue Feb 14 03:51:02 2017] tp_osd_tp cpuset=/ mems_allowed=0-1

>> > >> [Tue Feb 14 03:51:02 2017] CPU: 20 PID: 11864 Comm: tp_osd_tp Not

>> > >> tainted

>> > >> 3.10.0-327.13.1.el7.x86_64 #1

>> > >> [Tue Feb 14 03:51:02 2017] Hardware name: HP ProLiant XL420

>> > >> Gen9/ProLiant

>> > >> XL420 Gen9, BIOS U19 09/12/2016

>> > >> [Tue Feb 14 03:51:02 2017]  ffff8819ccd7a280 0000000030e84036

>> > >> ffff881fa58f7528 ffffffff816356f4

>> > >> [Tue Feb 14 03:51:02 2017]  ffff881fa58f75b8 ffffffff8163068f

>> > >> ffff881fa3478360 ffff881fa3478378

>> > >> [Tue Feb 14 03:51:02 2017]  ffff881fa58f75e8 ffff8819ccd7a280

>> > >> 0000000000000001 000000000001f65f

>> > >> [Tue Feb 14 03:51:02 2017] Call Trace:

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff816356f4>] dump_stack+0x19/0x1b

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8163068f>]

>> > >> dump_header+0x8e/0x214

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8116ce7e>]

>> > >> oom_kill_process+0x24e/0x3b0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8116c9e6>] ?

>> > >> find_lock_task_mm+0x56/0xc0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8116d6a6>]

>> > >> *out_of_memory+0x4b6/0x4f0*

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81173885>]

>> > >> __alloc_pages_nodemask+0xa95/0xb90

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811b792a>]

>> > >> alloc_pages_vma+0x9a/0x140

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811976c5>]

>> > >> handle_mm_fault+0xb85/0xf50

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811957fb>] ?

>> > >> follow_page_mask+0xbb/0x5c0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81197c2b>]

>> > >> __get_user_pages+0x19b/0x640

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8119843d>]

>> > >> get_user_pages_unlocked+0x15d/0x1f0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8106544f>]

>> > >> get_user_pages_fast+0x9f/0x1a0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8121de78>]

>> > >> do_blockdev_direct_IO+0x1a78/0x2610

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81218c40>] ? I_BDEV+0x10/0x10

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8121ea65>]

>> > >> __blockdev_direct_IO+0x55/0x60

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81218c40>] ? I_BDEV+0x10/0x10

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81219297>]

>> > >> blkdev_direct_IO+0x57/0x60

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81218c40>] ? I_BDEV+0x10/0x10

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8116af63>]

>> > >> generic_file_aio_read+0x6d3/0x750

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffffa038ad5c>] ?

>> > >> xfs_iunlock+0x11c/0x130 [xfs]

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811690db>] ?

>> > >> unlock_page+0x2b/0x30

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81192f21>] ?

>> > >> __do_fault+0x401/0x510

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff8121970c>]

>> > >> blkdev_aio_read+0x4c/0x70

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811ddcfd>]

>> > >> do_sync_read+0x8d/0xd0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811de45c>] vfs_read+0x9c/0x170

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff811df182>]

>> > >> SyS_pread64+0x92/0xc0

>> > >> [Tue Feb 14 03:51:02 2017]  [<ffffffff81645e89>]

>> > >> system_call_fastpath+0x16/0x1b

>> > >>

>> > >>

>> > >> Feb 14 03:51:40 fr-paris kernel: *Out of memory: Kill process 7657

>> > >> (ceph-osd) score 45 or sacrifice child*

>> > >> Feb 14 03:51:40 fr-paris kernel: Killed process 7657 (ceph-osd)

>> > >> total-vm:8650208kB, anon-rss:6124660kB, file-rss:1560kB

>> > >> Feb 14 03:51:41 fr-paris systemd:* ceph-osd@3.service: main process

>> > >> exited, code=killed, status=9/KILL*

>> > >> Feb 14 03:51:41 fr-paris systemd: Unit ceph-osd@3.service entered

>> > >> failed

>> > >> state.

>> > >> Feb 14 03:51:41 fr-paris systemd: *ceph-osd@3.service failed.*

>> > >> Feb 14 03:51:41 fr-paris systemd: cassandra.service: main process

>> > >> exited,

>> > >> code=killed, status=9/KILL

>> > >> Feb 14 03:51:41 fr-paris systemd: Unit cassandra.service entered

>> > >> failed

>> > >> state.

>> > >> Feb 14 03:51:41 fr-paris systemd: cassandra.service failed.

>> > >> Feb 14 03:51:41 fr-paris ceph-mgr: 2017-02-14 03:51:41.978878

>> > >> 7f51a3154700 -1 mgr ms_dispatch osd_map(7517..7517 src has

>> > >> 6951..7517) v3

>> > >> Feb 14 03:51:42 fr-paris systemd: Device

>> > >> dev-disk-by\x2dpartlabel-ceph\x5cx20block.device

>> > >> appeared twice with different sysfs paths

>> > >> /sys/devices/pci0000:00/0000:0

>> > >> 0:03.2/0000:03:00.0/host0/target0:0:0/0:0:0:9/block/sdj/sdj2 and

>> > >> /sys/devices/pci0000:00/0000:00:03.2/0000:03:00.0/host0/targ

>> > >> et0:0:0/0:0:0:4/block/sde/sde2

>> > >> Feb 14 03:51:42 fr-paris ceph-mgr: 2017-02-14 03:51:42.992477

>> > >> 7f51a3154700 -1 mgr ms_dispatch osd_map(7518..7518 src has

>> > >> 6951..7518) v3

>> > >> Feb 14 03:51:43 fr-paris ceph-mgr: 2017-02-14 03:51:43.508990

>> > >> 7f51a3154700 -1 mgr ms_dispatch mgrdigest v1

>> > >> Feb 14 03:51:48 fr-paris ceph-mgr: 2017-02-14 03:51:48.508970

>> > >> 7f51a3154700 -1 mgr ms_dispatch mgrdigest v1

>> > >> Feb 14 03:51:53 fr-paris ceph-mgr: 2017-02-14 03:51:53.509592

>> > >> 7f51a3154700 -1 mgr ms_dispatch mgrdigest v1

>> > >> Feb 14 03:51:58 fr-paris ceph-mgr: 2017-02-14 03:51:58.509936

>> > >> 7f51a3154700 -1 mgr ms_dispatch mgrdigest v1

>> > >> Feb 14 03:52:01 fr-paris systemd: ceph-osd@3.service holdoff time

>> > >> over,

>> > >> scheduling restart.

>> > >> Feb 14 03:52:02 fr-paris systemd: *Starting Ceph object storage

>> > >> daemon

>> > >> osd.3.*..

>> > >> Feb 14 03:52:02 fr-paris systemd: Started Ceph object storage daemon

>> > >> osd.3.

>> > >> Feb 14 03:52:02 fr-paris numactl: 2017-02-14 03:52:02.307106

>> > >> 7f1e499bb940

>> > >> -1 WARNING: the following dangerous and experimental features are

>> > >> enabled:

>> > >> bluestore,rocksdb

>> > >> Feb 14 03:52:02 fr-paris numactl: 2017-02-14 03:52:02.317687

>> > >> 7f1e499bb940

>> > >> -1 WARNING: the following dangerous and experimental features are

>> > >> enabled:

>> > >> bluestore,rocksdb

>> > >> Feb 14 03:52:02 fr-paris numactl: starting osd.3 at - osd_data

>> > >> /var/lib/ceph/osd/ceph-3 /var/lib/ceph/osd/ceph-3/journal

>> > >> Feb 14 03:52:02 fr-paris numactl: 2017-02-14 03:52:02.333522

>> > >> 7f1e499bb940

>> > >> -1 WARNING: experimental feature 'bluestore' is enabled

>> > >> Feb 14 03:52:02 fr-paris numactl: Please be aware that this feature

>> > >> is

>> > >> experimental, untested,

>> > >> Feb 14 03:52:02 fr-paris numactl: unsupported, and may result in data

>> > >> corruption, data loss,

>> > >> Feb 14 03:52:02 fr-paris numactl: and/or irreparable damage to your

>> > >> cluster.  Do not use

>> > >> Feb 14 03:52:02 fr-paris numactl: feature with important data.

>> > >>

>> > >> This seems to happen only in 11.2.0 and not in 11.1.x . Could you

>> > >> please

>> > >> help us in resolving this issue by means of any config change to

>> > >> limit the

>> > >> memory use on ceph-osd or a bug in the current kraken release.

>> > >>

>> > >> Thanks,

>> > >> Muthu

>> > >>

>> > >> _______________________________________________

>> > >> ceph-users mailing list

>> > >> ceph-users@xxxxxxxxxxxxxx

>> > >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

>> > >>

>> > >>

>> > >

>> > >

>> > > --

>> > > С уважением / Best regards

>> > >

>> > > Илья Летковский / Ilya Letkouski

>> > >

>> > > Phone, Viber: +375 29 3237335

>> > >

>> > > Minsk, Belarus (GMT+3)

>> > >

>> > _______________________________________________

>> > ceph-users mailing list

>> > ceph-users@xxxxxxxxxxxxxx

>> > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

>

>

>

> _______________________________________________

> ceph-users mailing list

> ceph-users@xxxxxxxxxxxxxx

> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

>

_______________________________________________

ceph-users mailing list

ceph-users@xxxxxxxxxxxxxx

http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

_______________________________________________

ceph-users mailing list

ceph-users@xxxxxxxxxxxxxx

http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

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