[Openstack-operators] Hypervisor free memory recommendations

Juan José Pavlik Salles jjpavlik at gmail.com
Thu Sep 4 18:25:59 UTC 2014


Yes, that's pretty much my scenario. The hosts are using about 1.x GB just
for being online (munin, nrpe, amanda, fail2ban, etc), so 10% in 16GB nodes
is a really bad idea. I'll stick to your proposal!

Thanks!

2014-09-04 15:18 GMT-03:00 Jay Pipes <jaypipes at gmail.com>:

> On 09/04/2014 01:51 PM, Juan José Pavlik Salles wrote:
>
>> Hi Jay, I do agree about 10% being too much memory in big nodes, but
>> right now we are using small ones (too small if you ask). These new
>> nodes are 16GB so if I reserve 4 Gb for the dom0 I'd be loosing 25% of
>> the available RAM. I was thinking about something like: if you have got
>> less than 32GB give 10% of it to the dom0 and If you have got more than
>> 32GB go with 4GB for the dom0. Maybe different environments will need
>> different rules, but this should work in most standar deployments I'd
>> say. Jay, you mentioned that big nodes running many VMs don't neet more
>> than 4GB of dedicated RAM, haven't you ever had any swapping situation
>> in that kind of scenarios?
>>
>
> One more thing of note... if you have a lot of custom host or IDS
> monitoring or custom host applications running, you may want to increase
> the host RAM. By "custom" I mean anything other than
> Nagios/Zabbix/Ganglia-type stuff, which tend to be pretty low-memory
> daemons...
>
> Do some investigation of a sample packed and sparse compute node and
> adjust as you see fit.
>
> Best,
> -jay
>



-- 
Pavlik Salles Juan José
Blog - http://viviendolared.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20140904/fac8d7e5/attachment.html>


More information about the OpenStack-operators mailing list