Red Hat Bugzilla – Bug 1360301
[RFE] allow qemu gfapi log redirection
Last modified: 2017-08-01 23:27:29 EDT
Description of problem: When qemu images are accessed via gfapi, troubleshooting is an issue as there are no logs unlike with fuse access. We require logging to be implemented for this path. Version-Release number of selected component (if applicable): NA How reproducible: NA
(In reply to Sahina Bose from comment #0) > Description of problem: > > When qemu images are accessed via gfapi, troubleshooting is an issue as > there are no logs unlike with fuse access. We require logging to be > implemented for this path. > Are you saying that there are no logs at all or whatever the log is there it is not sufficient? It would be good to have some more clarification on this bug.
No log file while accessing qemu via gfapi currently Prasanna, can you add more details to this bug?
Sure sahina :) I have put about improvement at least which I know and worked on and in qemu block layer, there were recent changes in Qemu that helps in choosing LOG LEVEL [1] and should land in RHEL-7.3. But with current code, the qemu gfapi logs fall on stderr and there is no way to configure them for redirection (internally) Patch [2] specifies an option for doing so, yet not part of qemu-2.7, since it was initiated at hard-freeze state, after a conversation maintainer confirms as it will not fall in bugs category couldn't make in 2.7, though it got a review. I hope it should go in 2.8. That's all for now. [1] https://www.mail-archive.com/qemu-devel@nongnu.org/msg364972.html [2] https://lists.gnu.org/archive/html/qemu-devel/2016-07/msg05439.html
Hi Rajesh, I have given a devel_ack on this, if you think this BZ needs to be moved out to a different component please do that and remove the ack and tracker. Thanks, Atin
read component as product in comment 4
For logging level, see Bug 1151859
We need the ability to redirect the gfapi logs to a log file instead of stderr to ensure supportability. From RHV 4.1, images on gluster volumes will be accessed via gfapi instead of fuse using the gluster driver in qemu. This path has no logging unlike the fuse access and troubleshooting customer issues will be a major problem.
Patch to support this is in QEMU 2.8.0: commit e9db8ff38e539260a2cb5a7918d1155b7d92a264 Author: Prasanna Kumar Kalever <prasanna.kalever@redhat.com> Date: Fri Jul 22 20:26:48 2016 +0530 block/gluster: add support to choose libgfapi logfile currently all the libgfapi logs defaults to '/dev/stderr' as it was hardcoded in a call to glfs logging api. When the debug level is chosen to DEBUG/TRACE, gfapi logs will be huge and fill/overflow the console view. This patch provides a commandline option to mention log file path which helps in logging to the specified file and also help in persisting the gfapi logs.
1. package version qemu-kvm-rhev-2.8.0-5.el7.x86_64 2. boot guest with -drive id=drive_image2,if=none,cache=none,format=qcow2,file=gluster://bootp-73-130-253.rhts.eng.pek2.redhat.com/gv0/stg1.qcow2,file.debug=1,file.logfile=/var/qemu-gfapi.log \ 3. get debug log from /var/qemu-gfapi.log 4. boot guest with file.debug=10,file.logfile=/var/qemu-gfapi.log \ I can get debug log as well, but the debug level should between 0~9
Created attachment 1259434 [details] file.debug=1
Created attachment 1259435 [details] file.debug=10
Hi Jeff, Could you check the test result, is it acceptable that debug log is generated when debug level is out of (0~9) Thanks Suqin
(In reply to Suqin Huang from comment #17) > Hi Jeff, > > Could you check the test result, is it acceptable that debug log is > generated when debug level is out of (0~9) > > Thanks > Suqin Yes, that is acceptable. Higher numbers are increasing verbosity; a level of 9 is the maximum level, so debug levels passed higher than 9 will be equivalent to 9.
Update to verified according to comment14 and comment18
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2017:2392