We are troubleshooting a performance issue where IQS freezes during bulk event processing. When a massive event occurs (such as large-scale OU movements), IQS generates an excessive amount of randomly named .txt files containing PS rule payloads. The system becomes unresponsive, requiring a hard restart, file cleanup, and temporary dissociation of the rules. Are there any known open bugs or documentation regarding this specific behavior?
@jsosa we’ve experienced a similar issue during one of our bulk assignment of access.
In our case, the root cause was that rules were triggered for every associated provisioning event. When a large number of provisioning events occur simultaneously, all available IQService resources are be consumed, causing the service to become unavailable and required restart.
If you have this type of activity frequently, increasing the server resources (CPU and memory) may help improve stability and performance.
Some additional practices that may help:
- Review your rule logic carefully and exit as early as possible when the execution conditions are not met. This reduces unnecessary processing.
- Be careful when using timers and avoid multiple processes reading from or writing to the same files simultaneously.
- If there are any planned bulk activities that do not require rule execution, consider temporarily disabling them during those operations.
Let’s see if others in the community have encountered this issue and can share additional recommendations or best practices.
Hi Ousmande thanks you for your response! In this case, we have an after modify that in the first line, it has a clause like
if operation != Disabled end
but iqs gets overloaded even when executed code is near 0.. so I think your suggestion of temporarily disable rules is the way.