Gary,
Two RGs only make sense if the services can move between nodes
independently. Given that they depend on one zpool and one IP address,
then the services can only be online on one node at any one time. As
such, putting them in the same RG will always lead to the problem you've
already encountered.
The only way to solve this is to give each service its own IP and its
own zpool.
Regards,
Tim
---
On 07/13/11 20:49, Gary Mills wrote:
The entity managed by the RGM is a Resource Group,
and all resources
within a single RG will be brought online/offline
together. The
intention is to group related resources so that they
are all
managed as a single unit.
That certainly seems reasonable. So I need one RG for
each data service.
If you have two unrelated services you should put
each in it's own
RG, You can then use the affinity and dependency
rules to manage
any interactions, i.e. determining start order.
They're unrelated in terms of the services they deliver, but
all of them are dependant on a single shared hostname and
a single zpool. How can I configure that? Start order of the
data services shouldn't matter, as long as they have the
necessary resources.
--
Tim Read
Software Developer
Solaris Availability Engineering
Oracle Corporation UK Ltd
Springfield
Linlithgow
EH49 7LR
Phone: +44 (0)1506 672 684
Mobile: +44 (0)7802 212 137
Twitter: @timread
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTICE: This email message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
_______________________________________________
ha-clusters-discuss mailing list
ha-clusters-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/ha-clusters-discuss
