[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-API] switches for XCP in High Avaibility

  • To: xen-api@xxxxxxxxxxxxx
  • From: George Shuklin <george.shuklin@xxxxxxxxx>
  • Date: Fri, 19 Jul 2013 19:14:07 +0400
  • Delivery-date: Fri, 19 Jul 2013 15:14:19 +0000
  • List-id: User and development list for XCP and XAPI <xen-api.lists.xen.org>

Is your iscsi target ready for highload? Check latency log during migration process.

There is a certain timeout for iscsi operations. If target could not serve data in timed manner, this cause errors.

On 18.07.2013 11:42, Joseph-Andre Guaragna wrote:

We came accros some network issues on our infrastructure when we
implemented a two node XCP with an ISCI target. We were using standard
switches (dlink), and got into troulble when our VMs (3) migrate from
one XCP to the other. The vms after a while were filling the XCP log
with I/O errors.

I wonder if any of you encountered this problems. A network told me to
increase mtu to 9k (jumbo frames) in order to reduce network traffic
and activate some kind of storm braodcast limitation( not stp ). If
somebody came accrossed this kind of problems, didi this this kind of
solutions solve their troubles of it was something else.

  I am probably to vague in my description, and can give anymore
detailled on our situation. I am stuck for a week now, woth a problem
arriving randomly (hevay load troublen, standard flawless).

Thanks for your help, Joseph

Xen-api mailing list

Xen-api mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.