Activity
From 10/25/2017 to 11/23/2017
11/23/2017
- 10:12 AM Bug #4233 (Resolved): krbd: verify use of all GFP_* flags
- All the low-hanging issues have been addressed, I believe -- GFP_ATOMIC calls, etc.
- 10:05 AM Bug #22005 (Need More Info): result -108 xferred 2000, blk_update_request: I/O error
- 10:03 AM Bug #20927: libceph does not give up reconnecting on broken network
- The existing "-o force" wouldn't help. The plan is to implement "-o full-force", hopefully for kernel 4.16.
OTOH,...
11/20/2017
- 03:43 AM Bug #20927: libceph does not give up reconnecting on broken network
- solved this problem ?
In the production environment, if net break,
manually execute commands (rbd unmap -o force ...
11/17/2017
- 08:32 AM Bug #22146: "mount error 22 = Invalid argument" in kcephfs-luminous-distro-basic-smithi
- kernel was not specified in these tests. I checked one log, test was run on 4.4 kernel. 4.4 kernel does not
underst...
11/16/2017
- 08:56 PM Bug #22146: "mount error 22 = Invalid argument" in kcephfs-luminous-distro-basic-smithi
- Also seems in this run
http://pulpito.ceph.com/yuriw-2017-11-16_16:18:53-knfs-luminous-distro-basic-smithi/ - 08:53 PM Bug #22146 (Rejected): "mount error 22 = Invalid argument" in kcephfs-luminous-distro-basic-smithi
- This is luminous v12.2.2 point release
Run: http://pulpito.ceph.com/yuriw-2017-11-16_16:19:21-kcephfs-luminous-dis...
11/15/2017
- 10:22 PM Bug #22134: "kernel: BUG: sleeping function called from invalid context at lib/ioremap.c" in knfs...
- Harmless warning, introduced in 4.14. I pulled in acpi updates into testing, it's now built -- reschedule!
- 04:35 PM Bug #22134: "kernel: BUG: sleeping function called from invalid context at lib/ioremap.c" in knfs...
- @Ilya can you take a look please?
- 04:26 PM Bug #22134 (Resolved): "kernel: BUG: sleeping function called from invalid context at lib/ioremap...
- This is luminous v12.2.2 point release
http://pulpito.ceph.com/yuriw-2017-11-14_21:08:56-kcephfs-luminous-testing-...
11/11/2017
- 09:36 AM Bug #22110: Failures in run_xfstests in krbd-luminous-testing-basic-smithi
- That's fine. Looks like it was a 12.2.2 run -- consider it approved!
- 09:32 AM Bug #22014 (Pending Backport): krbd: memory allocation related deadlock (cloned image)
- https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1e37f2f84680fa7f8394fd444b6928e334495cc...
11/10/2017
- 09:48 PM Bug #22110: Failures in run_xfstests in krbd-luminous-testing-basic-smithi
- @Ilya can you pls take a look ?
- 09:48 PM Bug #22110 (New): Failures in run_xfstests in krbd-luminous-testing-basic-smithi
- This is luminous v12.2.2 point release
Run: http://pulpito.ceph.com/yuriw-2017-11-09_17:14:04-krbd-luminous-testin...
11/09/2017
- 08:41 AM Bug #22014: krbd: memory allocation related deadlock (cloned image)
- [PATCH] rbd: use GFP_NOIO for parent stat and data requests
11/06/2017
- 04:22 PM Bug #21326 (Closed): xenial/4.4.0-93-generic rbd: sysfs write failed map failed: (110) Connection...
- 04:20 PM Bug #22005: result -108 xferred 2000, blk_update_request: I/O error
- Hi Alexey,
This snippet suggests that the ceph client instance on node2 got blacklisted.
Can you provide the e... - 10:28 AM Bug #22014 (In Progress): krbd: memory allocation related deadlock (cloned image)
11/03/2017
- 10:34 AM Bug #22014: krbd: memory allocation related deadlock (cloned image)
- This particular one is easy to fix, I'll patch it next week.
- 06:45 AM Bug #22014: krbd: memory allocation related deadlock (cloned image)
- happened on layered image
- 06:42 AM Bug #22014 (Resolved): krbd: memory allocation related deadlock (cloned image)
- calltrace of ceph-msgr
[7507093.943117] [<ffffffffa075dfc9>] jbd2_log_wait_commit+0xa9/0x130 [jbd2]
[7507093.94314...
11/02/2017
- 12:58 PM Bug #20998 (Resolved): RHEL74 GA kernel paniced on client node running smallfile tests with 3 act...
- 12:17 PM Bug #22005 (Need More Info): result -108 xferred 2000, blk_update_request: I/O error
- Hello, community.
RBD linux kernel client hangs, when cluster in rebalance process.
Cluster has 3 client nodes, ...
10/31/2017
- 03:54 PM Bug #21420: ceph_osdc_writepages(): pre-allocated osdc->msgpool_op messages vs large number of sn...
- 400 is correct. 4k is 4096 bytes -- the size of the pre-allocated buffer.
I've updated the docs in https://github... - 03:49 PM Bug #21420: ceph_osdc_writepages(): pre-allocated osdc->msgpool_op messages vs large number of sn...
- What is the correct upper bound on the number of snapshots - is it 400 or 4000? In the original description, you men...
10/30/2017
- 02:05 AM Feature #21944: kclient: implement copy_file_range
- I think even kernel VFS does not provide hook for this. kernel uses both splice_read and splice_write callbacks to im...
10/27/2017
- 09:03 AM Feature #21944 (Resolved): kclient: implement copy_file_range
- http://man7.org/linux/man-pages/man2/copy_file_range.2.html
It’s an in-kernel (or even served-side) copy of data b... - 07:10 AM Bug #21420: ceph_osdc_writepages(): pre-allocated osdc->msgpool_op messages vs large number of sn...
- No, not yet. Increasing the size of pre-allocated messages would be a one-line patch, but seems a bit dumb -- it's s...
10/26/2017
- 05:10 PM Bug #21420: ceph_osdc_writepages(): pre-allocated osdc->msgpool_op messages vs large number of sn...
- We are seeing this on other systems now that have fewer snapshots, it appears to happen around the time that the snap...
Also available in: Atom