Welcome to the new Gigaspaces XAP forum. To recover your account, please follow these instructions.

Ask Your Question
0

In Primary-Backup cluster, should I use PlatformCapability or Zones?

We are running in a cluster with one Primary machine and one Replicant(Backup) machine. Prior to GS7 we were using an SLA and PlatformCapability to ensure that the primary-space only ran on the Primary machine, and the backup-space only ran on the Replicant machine.

In GS7, should we continue to use PlatformCapability? If so, now that we are using gs-agent to start our GSC(s), what is the correct way to pass the capability xml file to gsc.sh at startup?

Or, is the 'zone' concept more appropriate for this type of configuration?

Thanks, Rowland

{quote}This thread was imported from the previous forum. For your reference, the original is [available here|http://forum.openspaces.org/thread.jspa?threadID=3246]{quote}

asked 2009-09-18 15:58:49 -0500

river2sea gravatar image

updated 2013-08-08 09:52:00 -0500

jaissefsfex gravatar image
edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
0

By default, with XAP 7.0 the even primary distribution option is turned on. This means that when deploying a Data-Grid, the GSM will try to provision primary and backups to different machines where it will try (as much as it can) to distribute the primary spaces across many machines as possible.
Directing backups to a specific GSC would not allow you to utilize all the machines in proper manner- you might end up with machines running only primary s[aces and others only backup spaces which means you will not make a good use of your HW.

Still , in some cases (when having a single partition with one primary space and one backup space) , you might want to specify as part of the deployment time the location/zone of each space.

The zone is the replacement of the "platform capability" we had with older versions. As you can see , it is much simpler to use.

Shay

answered 2009-09-18 16:09:25 -0500

shay hassidim gravatar image
edit flag offensive delete link more

Comments

Thanks Shay, I think we're getting closer. For our purposes, at least in v1 of our system, we want to specify exactly which machine has the primary space and exactly which machine has the backup space. I see how to specify which zone(s) to deploy a PU to on this page: http://www.gigaspaces.com/wiki/displa...

Now my question is - how do I specify what zone a machine belongs to when I start the gs-agent.sh?

Thanks, Rowland

river2sea gravatar imageriver2sea ( 2009-09-18 16:19:51 -0500 )edit

You should set the following system property via the EXT_JAVA_OPTIONS variable: {code}export EXT_JAVA_OPTIONS=-Dcom.gs.zones=myzone {code}

You do that by wraping the gsc script with your your script which set this variable and later calls the original gsc script.

You should configure the GS-Agent to call your scripts instead the original ones.

See: \gigaspaces-xap-premium-7.0.1-ga\config\gsa\gsc.xml {code} windows="${com.gs.home}/bin/gsc.bat" unix="${com.gs.home}/bin/gsc.sh"> {code} should be replaced with: {code} windows="mygsc.bat" unix="mygsc.sh"> {code}

Shay

shay hassidim gravatar imageshay hassidim ( 2009-09-18 16:29:32 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower

Stats

Asked: 2009-09-18 15:58:49 -0500

Seen: 75 times

Last updated: Sep 18 '09