Greg Zartman via smartos-discuss wrote:
On Fri, Feb 27, 2015 at 11:57 AM, Greg Zartman
<[email protected] <mailto:[email protected]>> wrote:
I was wondering if there was something going on with the way QEMU
does more of a round robin approach to providing resources to the
KVM instances. In a practical sense, the user wouldn't notice the
very small period of time with the VM wasn't being given access to
resources, but maybe the NFS service does. I was wondering if NFS
might somehow be timing out when it senses the kvm instance not
responding and is doing some sort or reconnect or error checking.
That's the way this is acting.
I'm convinced now that there is a problem with mounting GZ NFS shares
inside of a KVM instance. I setup a few test VMs with delegated
datasets, signal volume allocation, and then mounting nfs from another
appliance and there are no problems with moving data round except for
the VM that is mounting the GZ NFS share. I don't have anything more
technical at this point support this, but am going to do a bit more
testing.
I'll have to do some digging.
I know that NFS mounting into zones was always discouraged on Solaris
(and loop-back mount was a more appropriate solution), whether than same
caveats apply with SmartOS would be worth knowing.
--
Ian.
-------------------------------------------
smartos-discuss
Archives: https://www.listbox.com/member/archive/184463/=now
RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00
Modify Your Subscription:
https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb
Powered by Listbox: http://www.listbox.com
