Bug 1003818
| Summary: | qemu refuse to migrate when host ram over committed : error while loading state for instance 0x0 of device 'ram' | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Xiaoqing Wei <xwei> |
| Component: | qemu-kvm | Assignee: | Juan Quintela <quintela> |
| Status: | CLOSED NOTABUG | QA Contact: | Virtualization Bugs <virt-bugs> |
| Severity: | high | Docs Contact: | |
| Priority: | high | ||
| Version: | 7.0 | CC: | hhuang, juzhang, michen, qzhang, rbalakri, shuang, virt-maint, xwei |
| Target Milestone: | rc | ||
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2015-09-09 17:59:29 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
It is not little RAM overcommit. You are running 2 4GB Guests + the host OS on the same host, that is quite a bit of overcommit. just do a "free" after launching the source guest. As if it works on RHEL6.6 host, it uses memory diferently (it appears differently enough to make this work). Notice from the free command that you have got without swap, and without swap you just have used too much RAM. Could you do a ps -auxm while migration is happening? It appears that there is a memory leak somewhere. (In reply to Juan Quintela from comment #3) > > Could you do a ps -auxm while migration is happening? It appears that there > is a memory leak somewhere. Will do, but currently that origin host is doing Windows timer device tests, I would setup testbed for this BZ later. Best Regards, Xiaoqing Wei. Hmmm, tried several rounds of migration, no luck to reproduce.... I searched my test run history, it was happen in a test loop which has hundreds of cases, I will re-submit one to see if reproducible now. Regards, Xiaoqing. It hasn't been reproducible since February, and reporter didn't send a new submission. |
Description of problem: qemu refuse to migrate when host ram over committed : error while loading state for instance 0x0 of device 'ram' Version-Release number of selected component (if applicable): kernel-3.10.0-12.el7.x86_64 qemu-kvm-1.5.3-2.el7.x86_64 spice-server-0.12.4-1.el7.x86_64 seabios-bin-1.7.2.2-2.el7.noarch seavgabios-bin-1.7.2.2-2.el7.noarch sgabios-bin-0.20110622svn-3.el7.noarch ipxe-roms-qemu-20130517-1.gitc4bce43.el7.noarch How reproducible: 8/10 Steps to Reproduce: 1.on a 8GB Ram host boot a 4GB guest(src), 2.then boot a incoming vm for migrate(dest) -incoming tcp:0:5200 3.on src: migrate -d tcp:0:5200 Actual results: dest: QEMU 1.5.3 monitor - type 'help' for more information (qemu) info status VM status: paused (inmigrate) (qemu) qemu: warning: error while loading state for instance 0x0 of device 'ram' load of migration failed Expected results: migrate success even host has little RAM over commitment. Additional info: 1) host has swap enabled [root@dhcp-8-232 ~]# free -m total used free shared buffers cached Mem: 7607 6283 1324 0 42 2106 -/+ buffers/cache: 4133 3474 Swap: 7951 7929 22 2) same test can pass on rhel6 host