Once upon a time, Yaniv Kaul <ykaul@redhat.com> said:
> On Tue, Apr 11, 2017 at 3:57 PM, Chris Adams <cma@cmadams.net> wrote:
> > I've been getting an occasional message like:
> >
> > Storage domain hosted_storage experienced a high latency of
> > 5.26121 seconds from host node3.
> >
> > I'm not sure what is causing them though. I look at my storage
> > (EqualLogic iSCSI SAN) and storage network switches and don't see any
> > issues.
> >
> > When the above message was logged, node3 was not hosting the engine
> > (doesn't even have engine HA installed), nor was it the SPM, so why
> > would it have even been accessing the hosted_storage domain?
> >
>
> All hosts are monitoring their access to all storage domains in the data
> center.
Okay. Is there any more information about what this message actually
means though? Is it read latency, write latency, a particular VM, etc.?
I can't find any issue at the network or SAN level, nor any load events
that correlate with the times oVirt logs the latency messages.
Ovirt is reading 4k from the metadata special volume every 10 secods. If the read takes more than 5 seconds, you will see this warning in engine event log.
Maybe your storage or the host was overloaded at that time (e.g. vm backup)?
Nir