Bug #3927
closedkrbd: I/O errors (ENXIO) during rbd/kernel.sh workunit
0%
Description
I'm seeing ENXIO errors at what I believe to the "rbd/kernel.sh
teuthology workunit while testing the new request code. I don't
believe I saw those previously. It does not result in a failed
test.
I think the new code might be doing a better job of noticing errors,
and if that's the case maybe this is just an error that was going
unreported before.
Or it may be an error that should have been resolved in a way
different from what the new code is doing.
Whatever the case, I'd like to make sure I understand the cause.
Updated by Alex Elder about 11 years ago
This is the relevant portion of the yaml file:
- workunit:
clients:
all:
- rbd/map-unmap.sh
- rbd/kernel.sh
Here's the signature of the error on the console:
[32167.500532] libceph: client5158 fsid f770875c-668c-4874-8d8d-423eddd0b22c
[32167.543598] libceph: mon1 10.214.131.17:6789 session established
[32167.557777] rbd2: unknown partition table
[32167.557958] rbd: rbd2: added with size 0x40000000
[32167.708446] libceph: client5069 fsid f770875c-668c-4874-8d8d-423eddd0b22c
[32167.751436] libceph: mon2 10.214.131.10:6790 session established
[32167.807236] rbd2: unknown partition table
[32167.844258] rbd: rbd2: added with size 0x40000000
[32167.990978] libceph: client5139 fsid f770875c-668c-4874-8d8d-423eddd0b22c
[32168.033934] libceph: mon0 10.214.131.10:6789 session established
[32168.123956] rbd2: unknown partition table
[32168.160967] rbd: rbd2: added with size 0x40000000
. . . I think this is where rbd/kernel.sh starts . . .[32174.258189] rbd2: unknown partition table
[32174.295050] rbd: rbd2: added with size 0x493e000
[32177.251128] rbd3: unknown partition table
[32177.287141] rbd: rbd3: added with size 0x493e000
[32183.905673] rbd: rbd3: obj_request read result -6
[32183.905673]
[32183.973376] end_request: I/O error, dev rbd3, sector 0
[32184.008985] Buffer I/O error on device rbd3, logical block 0
[32184.045172] Buffer I/O error on device rbd3, logical block 1
[32184.080658] Buffer I/O error on device rbd3, logical block 2
[32184.114962] Buffer I/O error on device rbd3, logical block 3
[32184.148401] rbd: rbd3: obj_request read result -6
[32184.148401]
[32184.208322] end_request: I/O error, dev rbd3, sector 0
[32184.240219] Buffer I/O error on device rbd3, logical block 0
Updated by Alex Elder about 11 years ago
Hey! I just looked at the test, and here's how it ends:
- remove snapshot and detect error from mapped snapshot
rbd snap rm --snap=snap1 testimg1
sudo dd if=/dev/rbd/rbd/testimg1@snap1 of=/tmp/img1.snap1 2>&1 | grep 'Input/output error'
So I think this is an excellent thing!
I would like Josh to verify this though.
Updated by Alex Elder about 11 years ago
- Subject changed from krbd: I/O errors (ENXIO) during to krbd: I/O errors (ENXIO) during rbd/kernel.sh workunit
Updated by Josh Durgin about 11 years ago
Yes, the ENXIO is expected. Assuming it's being propagated out to dd, and the test passes (outputs OK at the end of kernel.sh), we can close this.
Updated by Alex Elder about 11 years ago
- Status changed from New to Closed
It turns out this new behavior is a good thing, we're just
reporting errors now where we apparently did not previously.
Closing this.