Hi all,

I have the following issue:
After emergency reboot of hypervisor node (i.e. reset via ipmi), system vms
got stuck in "starting state":
http://thesuki.org/temp/ss/SS-20130416154808.png

When I look onto console through vnc I see either prompt for fsck or errors
related to r/o mounted root fs:
http://thesuki.org/temp/ss/SS-20130416155009.png

There is no way to destroy vm either from web interface or api
(cloudmonkey):

 cloudmonkey> destroy systemvm id=eb3adb37-96d1-4785-884d-5958526b75fc
Async query failed for jobid 3c2fd0d7-a0a0-427a-ae3e-dda96b7cf9e0
Error 530 We cannot stop VM[ConsoleProxy|v-1058-VM] when it is in state
Starting
 cloudmonkey>

I have to manually update state in mysql before I am able to destroy this
vm:

mysql> update vm_instance set state='Stopped' where name='v-1058-VM';
\Query OK, 1 row affected (0.08 sec)
Rows matched: 1  Changed: 1  Warnings: 0

 cloudmonkey> destroy systemvm id=eb3adb37-96d1-4785-884d-5958526b75fc
.



....


...............



................

..............................................................





accountid =
b080d9c2-c3b2-4ffd-8a1c-04e99d03f5ef.......................................................
cmd = com.cloud.api.commands.DestroySystemVmCmd
created = 2013-04-16T14:54:47+0200
jobid = 23942741-a26c-4dba-b65b-1a4536aab1d7
jobprocstatus = 0
jobresult:
=========
systemvm:
========
id = eb3adb37-96d1-4785-884d-5958526b75fc
created = 2013-04-16T13:02:03+0200
dns1 = 8.8.8.8
hostid = b069e873-675b-435d-a7f1-a7edf87ce9b1
hostname = ad112.colobridge.net
name = v-1058-VM
objectId = 1058
 podid = f04bddae-fd73-47ac-b40e-32a4d2a8b7d5
state = Stopped
systemvmtype = consoleproxy
templateid = e54cbb23-bc5b-48d1-925b-1225ef938ea4
zoneid = 6ff79c9f-eccc-466f-9d43-3ce941855002
zonename = Zone1
jobresultcode = 0
jobresulttype = object
jobstatus = 1
userid = 79c285d1-54df-4169-84d1-72ff99f52998


First question is: why cloudstack does not handle such vm start errors ?

Second question is: how to force deletion of such VMs without mysql hack ?

environment:
CS 4.0.1, Centos 6.4 (management+node1+node2), OpenIndiana NFS server as
primary and secondary storage
-- 
Regards,
Valery

http://protocol.by/slayer

Reply via email to