I'm really sorry everyone, I'm sending the message below again afresh, to get 
it out of the thread I originally crafted my message from. (Note to self: don't 
start a new thread by replying to a message within an old one.  Head slap)


On Nov 20, 2009, at 12:42 PM, Jonathan Loran wrote:

> 
> Hi Hung-Sheng,
> 
> In discussing my problem with various engineers, this mistake keeps coming 
> up.  We plan on running the Sunray server software in the global zones of 
> each cluster member, in a non-clustered configuration.  We only want to 
> cluster the Solaris 10 branded zones/containers, not the Sunray server 
> software.  Sorry if I hadn't explained that clearly.
> 
> Thanks,
> 
> Jon
> 
> On Nov 19, 2009, at 3:55 PM, Dr. Hung-Sheng Tsao wrote:
> 
>> 
>> 
>> Jonathan Loran wrote:
>>> 
>>> Hi List,
>>> 
>>> I am seeking advice as to the feasibility of a project I'm working on:
>>> 
>>> Background:  My sysadmin group supports a satellite missions ops center 
>>> that relies on software which only works on Solaris SPARC systems.  The ops 
>>> center currently has 22 Sunblade 1500's, one for each satellite used for 
>>> data acquisition and commanding.  The overall system is networked such that 
>>> we have central name service, and storage, yet the actual workstations are 
>>> used to control missions and acquire data.  My plan is to replace each of 
>>> these workstations with Solaris 10 branded containers, running on two 
>>> T5120's setup in an HA configuration.  Each container would use a Sunray as 
>>> it's head for the satellite operators.  There are many advantages to this 
>>> setup, but it is critical that the containers reside in an HA environment.
>> sunray server support multi server env no other ha servers is required
>> 
>>> 
>>> What I would love to do is use Open HA to build the cluster, using the 
>>> hardware minimized setup, so we don't need separate quorum hardware.  We 
>>> have limited funds. We already own the T5120's and some Sunrays, but don't 
>>> have the cash for expensive shared storage and controllers.   As it stands 
>>> right now, we can't really build out a test environment, because the 
>>> Solaris 10 branded zones don't work on snv_111b, and as I understand it, 
>>> the current Open HA release won't work on the newer builds.  So my hope is 
>>> that there will be an Open HA release coincident with Opensolaris 2010.02.  
>>> I also hope there will be a Solaris 10 branded zone that is compatible in 
>>> 2010.02.  Is this a reasonable expectation?
>>> 
>>> Of course, there are other problems with this setup:
>>> 
>>> System recovery scenarios:   Since there's no boot CD for SPARC, recovering 
>>> a system could be tricky.  We will need a boot server within our operations 
>>> center to recover these systems of they completely die.  OTOH, hopefully 
>>> there will be a SPARC boot CD for 2010.02.
>>> 
>>> System updates:  These will need to be applied with extreme care, since we 
>>> don't want to break the cluster.  Unlike the current setup, all of our eggs 
>>> will be in one basket on the Sunray/zone servers.
>>> 
>>> What am I not thinking about?
>>> 
>>> Thanks in advance for any thoughts/recommendations on this.
>>> 
>>> Jon
>>> 
>>> 
>>> -     _____/     _____/      /           - Jonathan Loran -           -
>>> -    /          /           /                IT Manager               -
>>> -  _____  /   _____  /     /     Space Sciences Laboratory, UC Berkeley
>>> -        /          /     /      (510) 643-5146 jloran at ssl.berkeley.edu 
>>> <mailto:jloran at ssl.berkeley.edu>
>>> - ______/    ______/    ______/           AST:7731^29u18e3
>>>                                 
>>> 
>>> 
>>> ------------------------------------------------------------------------
>>> 
>>> _______________________________________________
>>> ha-clusters-discuss mailing list
>>> ha-clusters-discuss at opensolaris.org
>>> http://mail.opensolaris.org/mailman/listinfo/ha-clusters-discuss
>>>  
>> <hung-sheng_tsao.vcf>
> 

Reply via email to