incubator-tashi-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Miha Stopar <miha.sto...@xlab.si>
Subject Re: Not able to get the nodemanager running
Date Wed, 06 Oct 2010 06:17:21 GMT
You can try "rm /var/tmp/VmControlQemu/*" on the host where node manager is running (if you
are running Qemu).

Cheers,
Miha


----- Original Message -----
From: "Richard Gass" <richard.gass@intel.com>
To: tashi-user@incubator.apache.org
Sent: Tuesday, October 5, 2010 8:17:37 PM
Subject: RE: Not able to get the nodemanager running

That is what I run too.

-----Original Message-----
From: Michael Stroucken [mailto:mxs@cmu.edu] 
Sent: Tuesday, October 05, 2010 12:52 PM
To: tashi-user@incubator.apache.org
Subject: Re: Not able to get the nodemanager running

Why are you running the pickle data store?

Greetings,
Michael.

Richard Gass wrote:
> Try this,
> 1. stop the nodemanager,
> 2. delete the file /var/tmp/nm.dat,
> 3.  stop any running vms
> 4.  on the clustermanager, do data.baseDataObject.instances[-1] and
> make sure there is nothing there
> 5.  Restart the nodemanager
>
> On Mon, Oct 4, 2010 at 5:33 PM, deepa diwakar <deepa.diwakar1@gmail.com> wrote:
>   
>> Hi Richard,
>> We tried out the workaround you suggested, but still get the same error.
>>
>> del data.baseDataObject.instances[-1]
>> returns -1
>> and even if we save the baseDataObject, when we start the nodemanager we hit
>> the same issue.
>> Is there any other way we could resolve this issue? Where does Tashi store
>> the instances info, so we can try to manually delete it?
>> Thanks!
>>
>> Regards,
>> Deepa and Shobha
>>
>> On Mon, Oct 4, 2010 at 6:40 PM, Richard Gass <richardgass@gmail.com> wrote:
>>     
>>> ahh, the "-1 instance".  I have complained about this one myself.  I
>>> haven't seen it in a while and cannot remember exactly what the proper
>>> procedure is to get rid of it.
>>>
>>> Go into the clustermanager session and type
>>> data.baseDataObject.instances[-1]
>>>
>>> This will print the info for that instance.  You can the add a "del"
>>> before the previous command to get rid of it.  After you delete, be
>>> sure to save the object to disk with data.baseDataObject.save()
>>>
>>>
>>> RIchard
>>>
>>>
>>> On Mon, Oct 4, 2010 at 1:43 PM, deepa diwakar <deepa.diwakar1@gmail.com>
>>> wrote:
>>>       
>>>> Hi Richard,
>>>>
>>>> I am getting the following error when I try to run the nodemanager-
>>>>  [__main__:INFO] Using configuration file(s) ['./etc/TashiDefaults.cfg',
>>>> './etc/NodeManager.cfg']
>>>> [/tashi/trunk/src/tashi/nodemanager/nodemanagerservice.py:WARNING]
>>>> vmcontrol
>>>> backend does not report 113
>>>>
>>>> and the clustermanager in debug mode shows the following errors
>>>>         
>>>>>>> 2010-10-04 13:35:51,977
>>>>>>> [tashi.clustermanager.clustermanagerservice:ERROR] Got vmUpdate
for
>>>>>>> unknown
>>>>>>> instanceId -1
>>>>>>>               
>>>> Traceback (most recent call last):
>>>>   File "/tashi/trunk/src/tashi/clustermanager/clustermanagerservice.py",
>>>> line 401, in vmUpdate
>>>>     oldInstance = self.data.acquireInstance(instanceId)
>>>>   File "/tashi/trunk/src/tashi/clustermanager/data/getentoverride.py",
>>>> line
>>>> 36, in acquireInstance
>>>>     return self.baseDataObject.acquireInstance(instanceId)
>>>>   File "/tashi/trunk/src/tashi/clustermanager/data/fromconfig.py", line
>>>> 95,
>>>> in acquireInstance
>>>>     raise TashiException(d={'errno':Errors.NoSuchInstanceId,'msg':"No
>>>> such
>>>> instanceId - %d" % (instanceId)})
>>>> TashiException: {'msg': 'No such instanceId - -1', 'errno': 2}
>>>>
>>>> I tried to create a new setup of tashi on the same machine but it still
>>>> did
>>>> not work for me. which file stores info about existing instanceIDs? so I
>>>> can
>>>> resovle the problem?Thanks for your help!!
>>>>
>>>> Regards,
>>>> Deepa and Shobha
>>>>
>>>>         
>>>
>>> --
>>> Richard Gass
>>>       
>>     
>
>
>
>   


Mime
View raw message