Resource created on single node

Hi,

In case of unexpected offline/fault/termination of resource, there are 2 ways to restart.

                Type::RestartLimit

Type level attribute RestartLimit defines whether VCS attempts to restart a failed resource before informing of the fault. If the RestartLimit attribute is set to a non-zero value, the agent attempts to restart the resource before declaring the resource as faulted.

This is very useful if user is skeptical about particular type of resources. Set this attribute at Type level and all resources of that type will restarted. With this feature, only faulted resource is restarted. Service Group and other dependent/independent resources aren’t affected.

E.g.

type.cf snippet...
 
type Application (
.
.
        static int RestartLimit = 3
.
.
)

hatype command...

# hatype -display Application -attribute RestartLimit
#Type        Attribute              Value
Application  RestartLimit           3

In above example, all the Application resources will be retried atleast thrice in case of unexpected offline/fault/termination.

                Group::OnlineRetryLimit

Service Group level attribute OnlineRetryLimit allows a service group to be brought online again on the same system if a non-persistent resource in the service group faults. If, for some reason, the service group cannot be restarted, the VCS engine repeatedly tries to bring the service group online till the number of attempts that are specified by OnlineRetryLimit expires.

This is very useful if user is skeptical about particular Service Group. With this feature, entire Service Group is restarted. Thus, its useful if user wants service group to be restarted while faulted resource is restarted.

main.cf snippet...

group test_sg (
.
.
        OnlineRetryLimit = 5
.
.
        )

hagrp command...

# hagrp -modify test_sg OnlineRetryLimit 5
# hagrp -display test_sg -attribute OnlineRetryLimit
#Group       Attribute             System                 Value
test_sg      OnlineRetryLimit      localclus              5

In the above example, test_sg service group will be retried atleast 5 times if any of its resource faults.

Thanks & Regards,
Sunil Y