Re: [Xen-devel] [libvirt test] 58119: regressions - FAIL

[Jim Fehlig <jfehlig@xxxxxxxx>]

Ian Campbell wrote:
> On Mon, 2015-06-08 at 04:37 +0000, osstest service user wrote:
>   
>> flight 58119 libvirt real [real]
>> http://logs.test-lab.xenproject.org/osstest/logs/58119/
>>
>> Regressions :-(
>>
>> Tests which did not succeed and are blocking,
>> including tests which could not be run:
>>     
>
> This has been failing for a while now, sorry for not brining it to your
> attention sooner.
>   

I've noticed all the failures, but unfortunately they are the same and
contain no info as to why qemu failed to start.  The issue was discussed
during the last OpenStack+Xen+libvirt meetup, since Anthony is seeing a
similar issue with the OpenStack CI.

I've attempted to reproduced locally with a script that continuously
runs concurrent create, shutdown, and destroy, but eventually run into
other issues after many thousand iterations:

 # xl list
Name             ID     Mem   VCPUs  State   Time(s)
Domain-0         0      1871  48     r-----  152599.8
(null)           14058  0     1      --ps-d      24.8
(null)           14060  0     4      --p--d      20.9

libvirt has finished all cleanup of these domains.  qemu processes and
all xenstore entries associated with these domains are gone.  Yet
something has a refcnt on them.  From q debug key:

(XEN) General information for domain 14058:
(XEN)     refcnt=1 dying=2 pause_count=2
(XEN)     nr_pages=2 xenheap_pages=0 shared_pages=0 paged_pages=0
dirty_cpus={} max_pages=131328
(XEN)     handle=039e9ee6-4a84-3055-4c81-8ba426ae2656 vm_assist=00000004

(XEN) General information for domain 14060:
(XEN)     refcnt=1 dying=2 pause_count=2
(XEN)     nr_pages=2 xenheap_pages=0 shared_pages=0 paged_pages=0
dirty_cpus={} max_pages=131328
(XEN)     handle=04f95218-76b6-60f1-87f1-b51795e3b6ae vm_assist=00000000
(XEN)     paging assistance: hap refcounts translate external

Further debug suggestions appreciated :-).

My setup is running a recent xen.git (e13013db) and libvirt.git
(efc68de5) master.  libvirt contains an additional patch to cleanup
virDomainObj ref counting in the libxl driver, which fixed the first
problem the test script encountered.  I sent the patch to Anthony to
test in his OpenStack CI setup.

Regards,
Jim


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel