You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@cloudstack.apache.org by Garith Dugmore <ga...@saao.ac.za> on 2014/11/04 16:33:23 UTC

Re: HA issue and Xen resets

Update:

I've reinstalled everything below just incase I missed something but still no luck with my test instance coming up on my second xen host once the network cable has been yanked. 

Any recommendations or points will be greatly appreciated.

Pulling my hair out here...

side note: This recipe here -> http://discussions.citrix.com/topic/347884-automatic-applying-patches/ works well for updating Xen hosts from the CLI.

----- Original Message -----
> From: "Garith Dugmore" <ga...@saao.ac.za>
> To: users@cloudstack.apache.org
> Sent: Wednesday, October 29, 2014 12:06:27 PM
> Subject: Re: HA issue and Xen resets
> 
> After my live migration issues have now been resolved (in another thread) I'm
> now busy testing the HA capabilities again.
> 
> Current situation:
> 
> * ACS 4.4.1
> * Centos 6.5
> * 2 Xen hosts running 6.2 with sp1 and hotfixes
> * Ubuntu 14.04 instance running built from an ubuntu server 14.04 iso and
> using this guide ->
> http://softlit.blogspot.com/2014/05/xenserver-62-and-ubuntu-1404-lts.html
> with guest additions/PV drivers installed
> * Each xen host has only one network cable plugged into it
> * Running a basic zone with my management server as the primary and secondary
> storage which each xen server can mount correctly
> * My instances boot up with 10.8.1.x/24 ip's which are not routable on my
> network currently (incase this has a bearing on anything). I have a temp ip
> on my laptop to test ping's etc
> * 'alert.wait' is set to 30 in global settings and cloudstack-management
> restarted
> * My instance is using an HA enabled template. ie. Under "Details" for that
> instance "HA Enabled" does show "Yes"
> * Both my xen hosts are in the same cluster, the same pod and the same zone
> * The only virtual router that was auto created during my installation (I
> guess of the instance) is using a Compute offering of "System Offering For
> Software Router" which under "Service offerings", "System Offerings" and
> then "System Offering For Software Router" does have "Offer HA" set to "Yes"
> 
> What I've now tried is check the "Host" that the instance is running on under
> "Details" and yank that network cable (the xen host now has no network
> cables plugged into it at all) and wait for the failover to happen. After 5
> minutes the instance is still not reachable (I kept a window open with ping
> running).
> 
> Here is the log for 5 minutes -> http://pastebin.com/w7AsBqzt
> 
> What could be the problem?
> 
> Cheers,
> --
> 
> Garith Dugmore
> South African Astronomical Observatory
> and Southern African Large Telescope
> 
> !DSPAM:5450bc7216026803114569!
> 
> 
> 

-- 

Garith Dugmore 
South African Astronomical Observatory 
and Southern African Large Telescope