Modifying timeouts for executing repurposing request instances
Once a repurposing instance starts to run, it cannot be stopped or canceled. Time-outs can prevent repurposing instances from hanging indefinitely should a problem occur during their execution. multicluster features three timeouts to address most situations.
- Repurposing script maximum runtime: This setting controls the execution timeout for the move-in and move-out scripts. It is defined, in minutes, in the multicluster management console ().
- Repurposing instance timeout: This is the timeout for a repurposing instance. It is not directly
configurable but is based on the configuration of the repurposing script maximum runtime, as
follows:
repurposing_instance_timeout = script_max_runtime x 2 + 30The repurposing instance timeout is triggered in the following cases:- If there are no messages from the multicluster source and target proxies for the duration of the timeout.
- If the target resource plan does not require a change and no status message is received from the source proxy for the duration of the timeout.
In both cases, the timeout will cause the repurposing instance to enter an error state.
- SMC_REPURPOSE_INSTANCE_TIMEOUT: This timeout is defined in smcm.xml file
with a default of 12 hours. It is triggered if either the source or target proxy is disconnected
from the primary cluster for
the timeout duration after the repurposing instance has started. The timer is further influenced by
the following factors:
- If both the source and target proxies are connected, the timer is suspended but will restart from its current time if either proxy disconnects again.
- If the target resource plan does not require a change, then only the time when the source proxy is disconnected is counted.
- If the source proxy completes its part of the processing, source proxy disconnects will no longer cause a timeout.
For preceding timeouts 2 and 3, once the repurposing instance completes processing, no timeout can occur.