RHEV: spice VMs get killed with 'Lost connection with qemu process'
Issue
- Virtual Machines get killed
- qemu-kvm process is aborted.
- In RHEV-M, the following message is displayed:
ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (DefaultQuartzScheduler_Worker-13) [35d48c33] Correlation ID: null, Call Stack: null, Custom Event ID: -1, Message: VM <VM NAME> is down with error. Exit message: Lost connection with qemu process.
- In
qemu.log
, the following can be seen:
*** glibc detected *** /usr/libexec/qemu-kvm: double free or corruption (!prev): 0x00007ff2a1f97040 ***
======= Backtrace: =========
/lib64/libc.so.6(+0x75f4e)[0x7ff29c3b8f4e]
/lib64/libc.so.6(+0x78cf0)[0x7ff29c3bbcf0]
/usr/lib64/libspice-server.so.1(+0x1b36c)[0x7ff29cb8c36c]
/usr/lib64/libspice-server.so.1(+0x22087)[0x7ff29cb93087]
/usr/libexec/qemu-kvm(main_loop_wait+0x396)[0x7ff29f8f8916]
/usr/libexec/qemu-kvm(kvm_main_loop+0x13a)[0x7ff29f91c28a]
/usr/libexec/qemu-kvm(main+0x35d7)[0x7ff29f8fd3e7]
/lib64/libc.so.6(__libc_start_main+0xfd)[0x7ff29c361d5d]
/usr/libexec/qemu-kvm(+0x89f49)[0x7ff29f8f4f49]
Environment
- Red Hat Enterprise Virtualization
- Red Hat Enterprise Virtualization Hosts or Hypervisors (RHEV-H) with:
- RHEL6: spice package version lower than spice-0.12.4-15.el7
- RHEL7: spice package version lower than spice-0.12.4-17.el7
- Virtual Machines using
spice
console.
- Red Hat Enterprise Virtualization Hosts or Hypervisors (RHEV-H) with:
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.