- ceph-users - lists.ceph.io

OSD: FAILED ceph_assert(clone_size.count(clone))
by Jake Grimmett 23 Mar '20

23 Mar '20

Dear All, We are having problems with a critical osd crashing on a Nautilus (14.2.8) cluster. This is a critical failure, as the osd is part of a pg that is otherwise "down+remapped" due to other osd's crashing; we were hoping the pg was going to repair itself, as there are plenty of free osds, but for some reason this pg never managed to get out of an undersized state. The osd starts OK, runs for a few minutes, then crashes with an assert, immediately after trying to backfill the pg that is "down+remapped" -7> 2020-03-23 15:28:15.368 7f15aeea8700 5 osd.287 pg_epoch: 35531 pg[5.750s2( v 35398'3381328 (35288'3378238,35398'3381328] local-lis/les=35530/35531 n=190408 ec=1821/1818 lis/c 35530/22903 les/c/f 35531/22917/0 35486/35530/35530) [234,354,304,388,125,25,427,226,77,154]/[2147483647,2147483647,287,388,125,25,427,226,77,154]p287(2) backfill=[234(0),304(2),354(1)] r=2 lpr=35530 pi=[22903,35530)/9 rops=1 crt=35398'3381328 lcod 0'0 mlcod 0'0 active+undersized+degraded+remapped+backfilling mbc={} trimq=112 ps=121] backfill_pos is 5:0ae00653:::1000e49a8c6.000000d3:head -6> 2020-03-23 15:28:15.381 7f15cc9ec700 10 monclient: get_auth_request con 0x555b2f229800 auth_method 0 -5> 2020-03-23 15:28:15.381 7f15b86bb700 2 osd.287 35531 ms_handle_reset con 0x555b2fef7400 session 0x555b2f363600 -4> 2020-03-23 15:28:15.391 7f15c04c5700 5 prioritycache tune_memory target: 4294967296 mapped: 805339136 unmapped: 1032192 heap: 806371328 old mem: 2845415832 new mem: 2845415832 -3> 2020-03-23 15:28:15.420 7f15cc9ec700 10 monclient: get_auth_request con 0x555b2fef7800 auth_method 0 -2> 2020-03-23 15:28:15.420 7f15b86bb700 2 osd.287 35531 ms_handle_reset con 0x555b2fef7c00 session 0x555b2f363c00 -1> 2020-03-23 15:28:15.476 7f15aeea8700 -1 /home/jenkins-build/build/workspace/ceph-build/ARCH/x86_64/AVAILABLE_ARCH/x86_64/AVAILABLE_DIST/centos7/DIST/centos7/MACHINE_SIZE/huge/release/14.2.8/rpm/el7/BUILD/ceph-14.2.8/src/osd/osd_types.cc: In function 'uint64_t SnapSet::get_clone_bytes(snapid_t) const' thread 7f15aeea8700 time 2020-03-23 15:28:15.470166 /home/jenkins-build/build/workspace/ceph-build/ARCH/x86_64/AVAILABLE_ARCH/x86_64/AVAILABLE_DIST/centos7/DIST/centos7/MACHINE_SIZE/huge/release/14.2.8/rpm/el7/BUILD/ceph-14.2.8/src/osd/osd_types.cc: 5443: FAILED ceph_assert(clone_size.count(clone)) osd log (127KB) is here: <https://www.mrc-lmb.cam.ac.uk/scicomp/ceph-osd.287.log.gz> /var/log/ceph/ceph-osd.287.log.gz when the osd was running, the pg state is as follows [root@ceph7 ~]# ceph pg dump | grep ^5.750 5.750 190408 0 804 190119 0 569643615603 0 0 3090 3090 active+undersized+degraded+remapped+backfill_wait 2020-03-23 14:37:57.582509 35398'3381328 35491:3265627 [234,354,304,388,125,25,427,226,77,154] 234 [NONE,NONE,287,388,125,25,427,226,77,154] 287 24471'3200829 2020-01-28 15:48:35.574934 24471'3200829 2020-01-28 15:48:35.574934 112 with the osd down: [root@ceph7 ~]# ceph pg dump | grep ^5.750 dumped all 5.750 190408 0 0 0 0 569643615603 0 0 3090 3090 down+remapped 2020-03-23 15:28:28.345176 35398'3381328 35532:3265613 [234,354,304,388,125,25,427,226,77,154] 234 [NONE,NONE,NONE,388,125,25,427,226,77,154] 388 24471'3200829 2020-01-28 15:48:35.574934 24471'3200829 2020-01-28 15:48:35.574934 This cluster is being used to backup a live cephfs cluster and has 1.8PB of data, including 30 days of snapshots. We are using 8+2 EC. Any help appreciated, Jake Note: I am working from home until further notice. For help, contact unixadmin(a)mrc-lmb.cam.ac.uk -- Dr Jake Grimmett Head Of Scientific Computing MRC Laboratory of Molecular Biology Francis Crick Avenue, Cambridge CB2 0QH, UK. Phone 01223 267019 Mobile 0776 9886539

1 0

can't get healthy cluster to trim osdmaps (13.2.8)
by Nikola Ciprich 23 Mar '20

23 Mar '20

Hello dear cephers, lately, there's been some discussion about slow requests hanging in "wait for new map" status. At least in my case, it's being caused by osdmaps not being properly trimmed. I tried all possible steps to force osdmap pruning (restarting mons, restarting everyging, poking crushmap), to no avail. Still all OSDs keep min osdmap version 1, while newest is 4734. Otherwise cluster is healthy, with no down OSDs, network communication works flawlessly, all seems to be fine. Just can't get old osdmaps to go away.. I's very small cluster and I've moved all production traffic elsewhere, so I'm free to investigate and debug, however I'm out of ideas on what to try or where to look. Any ideas somebody please? The cluster is running 13.2.8 I'd be very grateful for any tips with best regards nikola ciprich -- ------------------------------------- Ing. Nikola CIPRICH LinuxBox.cz, s.r.o. 28.rijna 168, 709 00 Ostrava tel.: +420 591 166 214 fax: +420 596 621 273 mobil: +420 777 093 799 www.linuxbox.cz mobil servis: +420 737 238 656 email servis: servis(a)linuxbox.cz -------------------------------------

2 4

How to recover/mount mirrored rbd image for file recovery
by Ml Ml 23 Mar '20

23 Mar '20

Hello, my goal is to back up a proxmox cluster with rbd-mirror for desaster recovery. Promoting/Demoting, etc.. works great. But how can i access a single file on the mirrored cluster? I tried: root@ceph01:~# rbd-nbd --read-only map cluster5-rbd/vm-114-disk-1 --cluster backup /dev/nbd1 But i get: root@ceph01:~# fdisk -l /dev/nbd1 fdisk: cannot open /dev/nbd1: Input/output error dmesg shows stuff like: [Thu Mar 19 09:29:55 2020] nbd1: unable to read partition table [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) Here is my state: root@ceph01:~# rbd --cluster backup mirror pool status cluster5-rbd --verbose health: OK images: 3 total 3 replaying vm-106-disk-0: global_id: 0bc18ee1-1749-4787-a45d-01c7e946ff06 state: up+replaying description: replaying, master_position=[object_number=3, tag_tid=2, entry_tid=3], mirror_position=[object_number=3, tag_tid=2, entry_tid=3], entries_behind_master=0 last_update: 2020-03-19 09:29:17 vm-114-disk-1: global_id: 2219ffa9-a4e0-4f89-b352-ff30b1ffe9b9 state: up+replaying description: replaying, master_position=[object_number=390, tag_tid=6, entry_tid=334290], mirror_position=[object_number=382, tag_tid=6, entry_tid=328526], entries_behind_master=5764 last_update: 2020-03-19 09:29:17 vm-115-disk-0: global_id: 2b0af493-14c1-4b10-b557-84928dc37dd1 state: up+replaying description: replaying, master_position=[object_number=72, tag_tid=1, entry_tid=67796], mirror_position=[object_number=72, tag_tid=1, entry_tid=67796], entries_behind_master=0 last_update: 2020-03-19 09:29:17 More dmesg stuff: [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:29:55 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: 95 callbacks suppressed [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 0 [Thu Mar 19 09:30:02 2020] buffer_io_error: 94 callbacks suppressed [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 0, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 1 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 1, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 2 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 2, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 3 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 3, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 4 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 4, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 5 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 5, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 6 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 6, async page read [Thu Mar 19 09:30:02 2020] block nbd1: Other side returned error (30) [Thu Mar 19 09:30:02 2020] blk_update_request: I/O error, dev nbd1, sector 7 [Thu Mar 19 09:30:02 2020] Buffer I/O error on dev nbd1, logical block 7, async page read Do i have to stop the replaying or how can i mount the image on the backup cluster? Thanks, Michael

3 6

Re: [15.1.1-rc] - "Module 'dashboard' has failed: ('pwdUpdateRequired',)"
by Gencer W. Genç 23 Mar '20

23 Mar '20

Hi Lenz, Yeah, I saw the PR and i still hit the issue today. In the meantime while @Volker investigates, is there a workaround to bring dashboard back? Or should I wait for @Volker's investigation? P.S.: I fopund this too: https://tracker.ceph.com/issues/44271 Thanks, Gencer. ======= REPLY ABOVE ======= Hi Gencer, On 2020-03-22 09:37, Gencer W. Genç wrote: ... Hmm, I thought we had fixed that bug by merging the following fix: https://github.com/ceph/ceph/pull/33513 @Volker, would you mind taking a look at this? Thanks in advance! Lenz -- SUSE Software Solutions Germany GmbH - Maxfeldstr. 5 - 90409 Nuernberg GF: Felix Imendörffer, HRB 36809 (AG Nürnberg)

2 1

How can I recover PGs in state 'unknown', where OSD location seems to be lost?
by Mark S. Holliman 23 Mar '20

23 Mar '20

Hi all, I have a large distributed ceph cluster that recently broke with all PGs housed at a single site getting marked as 'unknown' after a run of the Ceph Ansible playbook (which was being used to expand the cluster at a third site). Is there a way to recover the location of PGs in this state, or a way to fall back to a previous config where things were working? Or a way to scan the OSDs to determine which PGs are housed there? All the OSDs are still in place and reporting as healthy, it's just the PG locations that are missing. For info: the ceph cluster is used to provide a single shared CephFS mount for a distributed batch cluster, and it includes workers and pools of OSDs from three different OpenStack clouds. Ceph version: 13.2.8 Here is the system health: [root@euclid-edi-ctrl-0 ~]# ceph -s cluster: id: 0fe7e967-ecd6-46d4-9f6b-224539073d3b health: HEALTH_WARN insufficient standby MDS daemons available 1 MDSs report slow metadata IOs Reduced data availability: 1024 pgs inactive 6 slow ops, oldest one blocked for 244669 sec, mon.euclid-edi-ctrl-0 has slow ops too few PGs per OSD (26 < min 30) services: mon: 4 daemons, quorum euclid-edi-ctrl-0,euclid-cam-proxy-0,euclid-imp-proxy-0,euclid-ral-proxy-0 mgr: euclid-edi-ctrl-0(active), standbys: euclid-imp-proxy-0, euclid-cam-proxy-0, euclid-ral-proxy-0 mds: cephfs-2/2/2 up {0=euclid-ral-proxy-0=up:active,1=euclid-cam-proxy-0=up:active} osd: 269 osds: 269 up, 269 in data: pools: 5 pools, 5120 pgs objects: 30.54 M objects, 771 GiB usage: 3.8 TiB used, 41 TiB / 45 TiB avail pgs: 20.000% pgs unknown 4095 active+clean 1024 unknown 1 active+clean+scrubbing OSD Pools: [root@euclid-edi-ctrl-0 ~]# ceph osd lspools 1 cephfs_data 2 cephfs_metadata 3 euclid_cam 4 euclid_ral 5 euclid_imp [root@euclid-edi-ctrl-0 ~]# ceph pg dump_pools_json dumped pools POOLID OBJECTS MISSING_ON_PRIMARY DEGRADED MISPLACED UNFOUND BYTES OMAP_BYTES* OMAP_KEYS* LOG DISK_LOG 5 0 0 0 0 0 0 0 0 0 0 1 16975540 0 0 0 0 79165311663 0 0 6243475 6243475 2 5171099 0 0 0 0 551991405 126879876 270829 3122183 3122183 3 8393436 0 0 0 0 748466429315 0 0 1556647 1556647 4 0 0 0 0 0 0 0 0 0 0 [root@euclid-edi-ctrl-0 ~]# ceph health detail ... PG_AVAILABILITY Reduced data availability: 1024 pgs inactive pg 4.3c8 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3ca is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3cb is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d0 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d1 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d2 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d3 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d4 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d5 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d6 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d7 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d8 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3d9 is stuck inactive for 246794.767182, current state unknown, last acting [] pg 4.3da is stuck inactive for 246794.767182, current state unknown, last acting [] ... [root@euclid-edi-ctrl-0 ~]# ceph pg map 4.3c8 osdmap e284992 pg 4.3c8 (4.3c8) -> up [] acting [] Cheers, Mark

1 0

Ceph pool quotas
by Stolte, Felix 23 Mar '20

23 Mar '20

Hey guys, a short question about pool quotas. Do they apply to stats attributes “stored” or “bytes_used” (Is replication counted for or not)? Regards Felix IT-Services Telefon 02461 61-9243 E-Mail: f.stolte(a)fz-juelich.de ------------------------------------------------------------------------------------- ------------------------------------------------------------------------------------- Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender), Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, Prof. Dr. Sebastian M. Schmidt ------------------------------------------------------------------------------------- -------------------------------------------------------------------------------------

3 2

Newbie to Ceph jacked up his monitor
by Jarett DeAngelis 23 Mar '20

23 Mar '20

hi folks. I seem to have gotten myself into a situation where my monitor directory filled itself up entirely while I wasn't looking, and now my monitor is unresponsive. (this is in a 3-node cluster). it grew a whole lot right after I added a number of OSDs and it started rebalancing. how can I get my monitor back into a working state so the rebalance can finish and the system becomes usable again? thx!

1 0

Maximum limit of lifecycle rule length
by Amit Ghadge 22 Mar '20

22 Mar '20

He All, We set rgw_lc_max_rules to 10000 but we seen the issue while the xml rule length are > 1MB It return InvalidRange, the format is below, <LifecycleConfiguration xmlns="http://s3.amazonaws.com/doc/2006-03-01/"> <Rule> <Status>Enabled</Status><Prefix>test_1/</Prefix><Expiration><Days>1</Days></Expiration>.</Rule> . . . </LifecycleConfiguration> Any reason why Ceph not allowed lc rule length > 1MB. Thanks, AmitG

1 0

Re: Cephfs mount error 1 = Operation not permitted
by Eugen Block 22 Mar '20

22 Mar '20

Hi, have you tried to mount with the secret only instead of a secret file? mount -t ceph ceph-n4:6789:/ /ceph -o name=client.1,secret=<SECRET> If that works your secret file is not right. If not you should check if the client actually has access to the cephfs pools ('ceph auth list'). Zitat von "Dungan, Scott A." <sdungan(a)caltech.edu>: > I am still very new to ceph and I have just set up my first small > test cluster. I have Cephfs enabled (named cephfs) and everything is > good in the dashboard. I added an authorized user key for cephfs with: > > ceph fs authorize cephfs client.1 / r / rw > > I then copied the key to a file with: > > ceph auth get-key client.1 > /tmp/client.1.secret > > Copied the file over to the client and then attempt mount witth the > kernel driver: > > mount -t ceph ceph-n4:6789:/ /ceph -o > name=client.1,secretfile=/root/client.1.secret > mount error 1 = Operation not permitted > > I looked in the logs on the mds (which is also the mgr and mon for > the cluster) and I don't see any events logged for this. I also > tried the mount command with verbose and I didn't get any further > detail. Any tips would be most appreciated. > > -- > > Scott Dungan > California Institute of Technology > Office: (626) 395-3170 > sdungan(a)caltech.edu<mailto:sdungan@caltech.edu> > > _______________________________________________ > ceph-users mailing list -- ceph-users(a)ceph.io > To unsubscribe send an email to ceph-users-leave(a)ceph.io

2 2

Cephfs mount error 1 = Operation not permitted
by Dungan, Scott A. 22 Mar '20

22 Mar '20

I am still very new to ceph and I have just set up my first small test cluster. I have Cephfs enabled (named cephfs) and everything is good in the dashboard. I added an authorized user key for cephfs with: ceph fs authorize cephfs client.1 / r / rw I then copied the key to a file with: ceph auth get-key client.1 > /tmp/client.1.secret Copied the file over to the client and then attempt mount witth the kernel driver: mount -t ceph ceph-n4:6789:/ /ceph -o name=client.1,secretfile=/root/client.1.secret mount error 1 = Operation not permitted I looked in the logs on the mds (which is also the mgr and mon for the cluster) and I don't see any events logged for this. I also tried the mount command with verbose and I didn't get any further detail. Any tips would be most appreciated. -- Scott Dungan California Institute of Technology Office: (626) 395-3170 sdungan(a)caltech.edu<mailto:sdungan@caltech.edu>

1 0