[one-users] Incorrect monitoring stats?

[sebastien goasguen]

So actually, I have the same problem with Xen and one 1.2.0....

I have to deploy by hand if FMEM is not enough.

-sebastien

On Wed, Jun 17, 2009 at 10:35 AM, Gordon Wells<gordon.wells at gmail.com> wrote:
> Hi
>
> I seem to have found a way around this, by running "xm mem-set 0 500"
> on the nodes FMEM seems to drop back to the original value after
> shutting down a VM. There doesn't seem to be load-balancing though,
> the first node gets filled with vms before they get run on the second
> (using either schedular). Is this correct?
>
> Regards
> --
>
>
> Gordon Wells
> Bioinformatics and Computational Biology Unit
> Department of Biochemistry
> University of Pretoria /
> High Performance Computing Research group
> Meraka Institute
> CSIR
>
> The first principle is that you must not fool yourself -- and you are
> the easiest person to fool. So you have to be very careful about that.
> After you've not fooled yourself, it's easy not to fool other
> scientists. You just have to be honest in a conventional way after
> that.
>
> -- Richard Feynman
>
>
>
>
> 2009/6/15 Gordon Wells <gordon.wells at gmail.com>:
>> Hi
>>
>> I'm using Scientific Linux 5.2 with xen. Will have a look at the
>> memory/ballooning situation.
>>
>> Regards
>>
>> --
>>
>>
>> Gordon Wells
>> Bioinformatics and Computational Biology Unit
>> Department of Biochemistry
>> University of Pretoria /
>> High Performance Computing Research group
>> Meraka Institute
>> CSIR
>>
>> The first principle is that you must not fool yourself -- and you are
>> the easiest person to fool. So you have to be very careful about that.
>> After you've not fooled yourself, it's easy not to fool other
>> scientists. You just have to be honest in a conventional way after
>> that.
>>
>> -- Richard Feynman
>>
>>
>>
>>
>> 2009/6/12 Ruben S. Montero <rubensm at dacya.ucm.es>:
>>> Hi Gordon
>>>
>>> With Xen, you can have problems if the Dom0 memory is not set to a
>>> "low" value. The information that can be obtained from xen varies from
>>> version to version so is difficult to get the available memory when
>>> using ballooning. If you are using KVM this should not happen and this
>>> is a bug, could you tell us the Linux version you are using...
>>>
>>> In the meantime, if you have enough memory for the VMs you can use the
>>> onevm deploy command to start VMs directly on a given host.
>>>
>>> Cheers
>>>
>>> Ruben
>>>
>>> On Tue, Jun 9, 2009 at 9:41 AM, Gordon Wells <gordon.wells at gmail.com> wrote:
>>>> Hi
>>>>
>>>> I seem to be getting the wrong info from "onehost list"
>>>>
>>>> TMEM = 4062208 and FMEM = 392192
>>>>
>>>> Yet /proc/meminfo says:
>>>>
>>>> MemTotal:      3593216 kB
>>>> MemFree:       3037364 kB
>>>> Buffers:         33088 kB
>>>> Cached:         288536 kB
>>>> SwapCached:          0 kB
>>>> Active:          73624 kB
>>>> Inactive:       279620 kB
>>>> HighTotal:           0 kB
>>>> HighFree:            0 kB
>>>> LowTotal:      3593216 kB
>>>> LowFree:       3037364 kB
>>>> SwapTotal:     2031608 kB
>>>> SwapFree:      2031608 kB
>>>> Dirty:             724 kB
>>>> Writeback:           0 kB
>>>> AnonPages:       31596 kB
>>>> Mapped:           8424 kB
>>>> Slab:            28020 kB
>>>> PageTables:       4208 kB
>>>> NFS_Unstable:        0 kB
>>>> Bounce:              0 kB
>>>> CommitLimit:   3828216 kB
>>>> Committed_AS:   225328 kB
>>>> VmallocTotal: 34359738367 kB
>>>> VmallocUsed:      2020 kB
>>>> VmallocChunk: 34359735827 kB
>>>>
>>>> As far as I can there is more than enough memory to run vms (1Gb
>>>> requested). These incorrect values seem to be preventing ON from doing
>>>> proper load balancing (or running the vms at all).
>>>>
>>>> Regards
>>>> Gordon
>>>> --
>>>>
>>>>
>>>> Gordon Wells
>>>> Bioinformatics and Computational Biology Unit
>>>> Department of Biochemistry
>>>> University of Pretoria /
>>>> High Performance Computing Research group
>>>> Meraka Institute
>>>> CSIR
>>>>
>>>> The first principle is that you must not fool yourself -- and you are
>>>> the easiest person to fool. So you have to be very careful about that.
>>>> After you've not fooled yourself, it's easy not to fool other
>>>> scientists. You just have to be honest in a conventional way after
>>>> that.
>>>>
>>>> -- Richard Feynman
>>>> _______________________________________________
>>>> Users mailing list
>>>> Users at lists.opennebula.org
>>>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>>>
>>>
>>>
>>>
>>> --
>>> +---------------------------------------------------------------+
>>>  Dr. Ruben Santiago Montero
>>>  Associate Professor
>>>  Distributed System Architecture Group (http://dsa-research.org)
>>>
>>>  URL:    http://dsa-research.org/doku.php?id=people:ruben
>>>  Weblog: http://blog.dsa-research.org/?author=7
>>>
>>>  GridWay, http://www.gridway.org
>>>  OpenNebula, http://www.opennebula.org
>>> +---------------------------------------------------------------+
>>>
>>
> _______________________________________________
> Users mailing list
> Users at lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>



-- 
---
Sebastien Goasguen
School of Computing
Clemson University
864-656-6753
http://runseb.googlepages.com