It's a full install from an XCP CD.

# service xapissl status
xapissl (pid  2323) is running...

I performed a restart of xapissl anyway, and xe-toolstack-restart still
fails starting the xapi service. Any other ideas? What would happen if I
were to restart the system or perform a reinstall?

Our customer and us are pretty nervous now, as they have not had a
successful backup since last week Friday. So what else can I try?

Thanks, Dave!

On 20 January 2012 15:41, Dave Scott <[email protected]> wrote:

> Hi,****
>
> ** **
>
> I should have asked earlier: is this a host installed via the XCP CD, or
> is this a Debian system running the xcp- packages?****
>
> ** **
>
> It looks like xapi can’t find a running stunnel, and it looks like the
> xe-toolstack-restart failed to run “/sbin/service xapissl restart”. Do you
> have stunnel running, and listening on port 443? On a system installed via
> the XCP CD, “service xapissl restart” should start stunnel. On a Debian
> system – I believe – the xapi init.d script itself starts stunnel.****
>
> ** **
>
> Dave****
>
> ** **
>
> *From:* Lars Seeliger [mailto:[email protected]]
> *Sent:* 20 January 2012 12:15
> *To:* Dave Scott
> *Cc:* [email protected]
> *Subject:* Re: [Xen-API] xapi will no longer start - what are my options?*
> ***
>
> ** **
>
> Hey, Dave
>
> Thanks for the prompt response. For fear of not including enough info,
> I've pasted the entire xensource.log contents here:
> http://pastebin.com/AW12gfM0
>
> If you need anything else, just shout; this problem has ruined my day! :p*
> ***
>
> On 20 January 2012 12:53, Dave Scott <[email protected]> wrote:****
>
> Hi Lars,****
>
>  ****
>
> Have a look in the main xapi logs (in the confusingly-named file
> /var/log/xensource.log). Start from the bottom and reverse-search to the
> string “XAPI SERVER STARTING”. The lines after that will show how far the
> startup sequence got.****
>
>  ****
>
> Cheers,****
>
> Dave****
>
>  ****
>
>  ****
>
>  ****
>
> *From:* [email protected] [mailto:
> [email protected]] *On Behalf Of *Lars Seeliger
> *Sent:* 20 January 2012 10:23
> *To:* [email protected]
> *Subject:* [Xen-API] xapi will no longer start - what are my options?****
>
>  ****
>
> Hi there
>
> A scripted backup running on one of our XCP installations failed a few
> days ago, while exporting a snapshot. After the failure I tried to delete
> the snapshot in question, to no avail (something about the VDI being in
> use).
>
> Anyway, I thought an xe-toolstack-restart would reset any lock on that
> file, allowing me to delete the no longer needed snapshot. Unfortunately,
> xapi is now unable to start.
>
> I've just tried again, and this appears in /var/log/messages:
>
> Jan 20 11:01:21 xcp-hoppe xapi: [ info|xcp-hoppe|0 thread_zero||watchdog]
> (Re)starting xapi...
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.850Z||1172|About to bind
> to /var/xapi/forker/fd_e8f89481-9aae-05a5-1d73-fbb713f58ea3
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.850Z||1172|bound,
> listening
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.851Z||2300|Child here!
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.851Z||2301|Grandchild
> here!
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.851Z||2301|Started:
> state.cmdargs = [/sbin/service;xapissl;restart]
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.851Z||2301|Started:
> state.env = [PATH=/sbin:/usr/sbin:/bin:/usr/bin]
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Selecting in
> handle_comms_no_fd_sock2
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Done
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|fd sock
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Selecting in
> handle_comms_with_fd_sock2
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Done
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|fd sock2
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Received fd
> named: ed350b47-3eb6-63e0-38c5-3beaaefb65dd - duping to 1 (from 6)
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Selecting in
> handle_comms_with_fd_sock2
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Done
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|fd sock2
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Received fd
> named: d810a903-961d-bb1e-aeb7-b39c98e5eefa - duping to 2 (from 6)
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Selecting in
> handle_comms_with_fd_sock2
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Done
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|comms sock
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Exec
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Finished...
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|Args after
> replacement = [/sbin/service;xapissl;restart]
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:25.852Z||2301|I've received
> the following fds: [2;1]
> Jan 20 11:04:26 xcp-hoppe fe: 20120120T10:04:26.942Z||2301|Caught
> unexpected exception: Unix.Unix_error(31, "write", "")
> Jan 20 10:05:41 xcp-hoppe HVM5652[12007]:    XENNET: WARNING: No handler
> for oid 0xffda0014.
> Jan 20 10:05:41 xcp-hoppe HVM5652[12007]:    XENNET: WARNING: No handler
> for oid 0xffa0ffa1.
> Jan 20 10:05:41 xcp-hoppe HVM5652[12007]:    XENNET: WARNING: Bad tcp task
> offload header?
> Jan 20 10:05:41 xcp-hoppe HVM5652[12007]:    XENNET: WARNING: Bad tcp task
> offload header?
> Jan 20 11:06:16 xcp-hoppe snmpd[15173]: Received SNMP packet(s) from UDP:
> [192.168.1.1]:2227
> Jan 20 11:06:16 xcp-hoppe snmpd[15173]: Received SNMP packet(s) from UDP:
> [192.168.1.1]:2228
> Jan 20 11:06:17 xcp-hoppe snmpd[15173]: Received SNMP packet(s) from UDP:
> [192.168.1.1]:2229
> Jan 20 11:06:17 xcp-hoppe snmpd[15173]: Received SNMP packet(s) from UDP:
> [192.168.1.1]:2230
> Jan 20 11:06:25 xcp-hoppe python: PERFMON: caught socket.error: (111
> Connection refused) - restarting XAPI session
> Jan 20 10:08:36 xcp-hoppe HVM5652[12007]: Time offset set 3569, added
> offset -1
> Jan 20 11:08:54 xcp-hoppe python: PERFMON: Caught signal 15 - exiting
> Jan 20 11:08:54 xcp-hoppe python: PERFMON: 11 Resource temporarily
> unavailable
> Jan 20 11:08:54 xcp-hoppe python: PERFMON: Traceback (most recent call
> last):
> Jan 20 11:08:54 xcp-hoppe python: PERFMON:   File
> "/opt/xensource/bin/perfmon", line 930, in ?     rc = main()
> Jan 20 11:08:54 xcp-hoppe python: PERFMON:   File
> "/opt/xensource/bin/perfmon", line 880, in main     cmd =
> cmdsock.recv(cmdmaxlen)
> Jan 20 11:08:54 xcp-hoppe python: PERFMON: error: (11, 'Resource
> temporarily unavailable')
> Jan 20 11:08:54 xcp-hoppe python: PERFMON: caught socket.error: (111
> Connection refused) - restarting XAPI session
> Jan 20 11:08:55 xcp-hoppe v6d: [ info|xcp-hoppe|0||watchdog] (Re)starting
> v6d...
> Jan 20 11:08:55 xcp-hoppe xapi: [ info|xcp-hoppe|0 thread_zero||watchdog]
> (Re)starting xapi...
> Jan 20 10:10:45 xcp-hoppe HVM5641[28792]: Time offset set 3563, added
> offset -1
> Jan 20 11:13:45 xcp-hoppe python: PERFMON: caught socket.error: (111
> Connection refused) - restarting XAPI session
>
>
> /var/log/SMI contains:
>
> [2598] 2012-01-20 11:08:55.277830    VASSR run
> ['/opt/xensource/sm/VASSR',
> '<methodCall><methodName>sr_get_driver_info</methodName><params><param><value><struct><member><name>host_ref</name><value>OpaqueRef:NULL</value></member><member><name>command</name><value>sr_get_driver_info</value></member><member><name>args</name><value><array><data/></array></value></member><member><name>device_config</name><value><struct/></value></member></struct></value></param></params></methodCall>']
> [2598] 2012-01-20 11:08:55.278332    Warning: vdi_[de]activate present
> for vastsky
> [2619] 2012-01-20 11:08:55.858537    Warning: vdi_[de]activate present
> for dummy
>
> Not sure there's anything of value in those logs...
>
> I'm somewhat desperate, as I'm unable to perform any xe commands and am
> worried a reboot of the server will not magically bring xapi back online,
> meaning the VMs will not start. This XCP host is critical to one of our
> customer's infrastructure. It's Friday and I could possibly go there this
> evening and perform tasks necessary to bring everything back online, I just
> don't quite know what my options are, aside from reboot and perhaps an
> in-place install of XCP.
>
> Does anyone have any bright ideas? I'm all ears!****
>
> ** **
>
_______________________________________________
xen-api mailing list
[email protected]
http://lists.xensource.com/mailman/listinfo/xen-api

Reply via email to