incubator-cloudstack-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nitin Mehta <Nitin.Me...@citrix.com>
Subject RE: Host in Alert state, how to remove it.
Date Mon, 21 May 2012 11:36:59 GMT
I agree. DB update is just a workaround and not a solution. Can you please file a bug at http://bugs.cloudstack.org
with your information and relevant logs for this ?

From: dan@soleks.com [mailto:dan@soleks.com]
Sent: Monday, May 21, 2012 12:01 PM
To: cloudstack-users@incubator.apache.org; Nitin Mehta
Cc: cloudstack-users@incubator.apache.org
Subject: RE: Host in Alert state, how to remove it.


I applied the following queries:

mysql> update host set status='Removed', removed=now() where id='13';

and

mysql> delete from host_details where host_id='13';

At least host is gone from UI, hope nothing is broken.

Technically it need to be fixed, because situation when host is not accessible and presented
in CS (in any state) very possible. I think enduser should be able to remove host from UI,
because doing so directly in the database you don't know what exactly need to be updated/removed.

> Hi Dan,
> I understand your concern but I think that this step should get you
> going. Since it's a soft delete it shouldn't be a problem.
> Let me know how it goes and if In some corner case this becomes an
> issue we will try to take care of that dependency as well because
> ultimately we want this host to be removed from the setup.
>
> -----Original Message-----
> From: dan@soleks.com<mailto:dan@soleks.com> [mailto:dan@soleks.com]<mailto:[mailto:dan@soleks.com]>
> Sent: Monday, May 21, 2012 11:18 AM
> To: cloudstack-users@incubator.apache.org<mailto:cloudstack-users@incubator.apache.org>
> Subject: RE: Host in Alert state, how to remove it.
>
>
>
>  Hi Nitin,
>
> I was thinking about it, but my concerns - it's direct databases
> intervention, and i afraid that not all dependences will be satisfied
> (not all relationship will be removed/updated).
>> Hi Dan,
>> I think you can just set the removed column for the particular host in
>> the db table host.
>>
>> Thanks,
>> -Nitin
>>
>> -----Original Message-----
>> From: dan@soleks.com<mailto:dan@soleks.com> [mailto:dan@soleks.com]<mailto:[mailto:dan@soleks.com]>
>> Sent: Sunday, May 20, 2012 10:39 PM
>> To: cloudstack-users@incubator.apache.org<mailto:cloudstack-users@incubator.apache.org>
>> Subject: Host in Alert state, how to remove it.
>>
>>
>>
>> Hi all,
>>
>> I got situation that host is in the Maintenance/Alert state, actual
>> physical host was reinstalled, so UUID gone. But i can't remove it
>> from CS (3.0.2).In the tomcat log im getting exception:
>>
>> <----------------cut-------------------->
>>
>> 2012-05-20 09:21:02,259 WARN  [cloud.cluster.ClusterManagerImpl]
>> (Cluster-Worker-89:null) Agent is unavailable
>> com.cloud.exception.AgentUnavailableException: Resource [Host:13] is
>> unreachable: Host 13: Host is not in the right state: Alert
>>        at
>> com.cloud.agent.manager.ClusteredAgentManagerImpl.getAttache(ClusteredAgentManagerImpl.java:494)
>>        at
>> com.cloud.agent.manager.AgentManagerImpl.send(AgentManagerImpl.java:494)
>>        at
>> com.cloud.agent.manager.AgentManagerImpl.send(AgentManagerImpl.java:1022)
>>        at
>> com.cloud.cluster.ClusterManagerImpl.sendToAgent(ClusterManagerImpl.java:448)
>>        at
>> com.cloud.cluster.ClusterManagerImpl.dispatchClusterServicePdu(ClusterManagerImpl.java:413)
>>        at
>> com.cloud.cluster.ClusterManagerImpl.access$500(ClusterManagerImpl.java:88)
>>        at
>> com.cloud.cluster.ClusterManagerImpl$3.run(ClusterManagerImpl.java:324)
>>        at
>> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110)
>>        at
>> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603)
>>        at java.lang.Thread.run(Thread.java:679)
>> 2012-05-20 09:21:02,260 DEBUG [cloud.cluster.ClusterManagerImpl]
>> (Cluster-Worker-5:null) Cluster PDU 95529533441 -> 95529533442.
>> agent: 0, pdu seq: 191, pdu ack seq: 190, json:
>> 2012-05-20 09:21:02,263 DEBUG
>> [cloud.cluster.ClusterServiceServletImpl] (Cluster-Worker-5:null) POST
>> http://192.168.230.146:9090/clusterservice response :true, responding
>> time: 3 ms
>> 2012-05-20 09:21:02,263 DEBUG [cloud.cluster.ClusterManagerImpl]
>> (Cluster-Worker-5:null) Cluster PDU 95529533441 -> 95529533442
>> completed. time: 3ms. agent: 0, pdu seq: 191, pdu ack seq: 190, json:
>>
>> <------------------cut---------------->
>>
>> Changing state to Up (directly in database), doesn't help:
>>
>> <----------------cut------------------>
>>
>> 2012-05-20 09:23:23,203 WARN  [cloud.cluster.ClusterManagerImpl]
>> (Cluster-Worker-90:null) Agent is unavailable
>> com.cloud.exception.AgentUnavailableException: Resource [Host:13] is
>> unreachable: Host 13: Host is not in the right state: Up
>>        at
>> com.cloud.agent.manager.ClusteredAgentManagerImpl.getAttache(ClusteredAgentManagerImpl.java:494)
>>        at
>> com.cloud.agent.manager.AgentManagerImpl.send(AgentManagerImpl.java:494)
>>        at
>> com.cloud.agent.manager.AgentManagerImpl.send(AgentManagerImpl.java:1022)
>>        at
>> com.cloud.cluster.ClusterManagerImpl.sendToAgent(ClusterManagerImpl.java:448)
>>        at
>> com.cloud.cluster.ClusterManagerImpl.dispatchClusterServicePdu(ClusterManagerImpl.java:413)
>>        at
>> com.cloud.cluster.ClusterManagerImpl.access$500(ClusterManagerImpl.java:88)
>>        at
>> com.cloud.cluster.ClusterManagerImpl$3.run(ClusterManagerImpl.java:324)
>>        at
>> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110)
>>        at
>> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603)
>>        at java.lang.Thread.run(Thread.java:679)
>> 2012-05-20 09:23:23,203 DEBUG [cloud.cluster.ClusterManagerImpl]
>> (Cluster-Worker-5:null) Cluster PDU 95529533441 -> 95529533442.
>> agent: 0, pdu seq: 193, pdu ack seq: 192, json:
>>
>> <----------------cut--------------->
>>
>> Seems like CS can't communicate with agent on the host, and not happy
>> about it. But host is gone, and it can't be
>> restored/rebooted/recovered.
>>
>> Dan/borei.
>>
>> ----------------------------------------------------------------
>> This message was sent using IMP, the Internet Messaging Program.
>>
>
>
> ----------------------------------------------------------------
> This message was sent using IMP, the Internet Messaging Program.
>


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message