PagerDuty Notifier CloudBlade Limits and Behavior
Lists all the PagerDuty Limits and Behaviors.
Where Can I Use This? | What Do I Need? |
|
- Prisma SD-WAN license
- PagerDuty Notifier CloudBlade
|
The PagerDuty Notifier CloudBlade operates on a fixed interval,
with a default of 300 seconds (5 minutes). During each cycle, it executes a specific
workflow to manage incidents. The CloudBlade also enforces limits to avoid placing
excessive load on the systems it interacts with.
SD-WAN Incidents (Open): In each cycle, the CloudBlade reads a
maximum of 5,000 open SD-WAN incidents. If more than 5,000 incidents are active,
the CloudBlade reads only the latest ones and ignores the rest.
- SD-WAN Incidents (Closed): For closed SD-WAN incidents, the CloudBlade reads
up to 1,000 incidents per cycle within a 3-hour sliding window.
PagerDuty Incidents (Created): To prevent overwhelming the
PagerDuty system, the CloudBlade can create a maximum of 1,000 new PagerDuty
incidents in each cycle. If there are 5,000 open SD-WAN incidents, it will take
approximately five cycles to forward them all.
- PagerDuty Incidents (Resolved): There is no hard limit on the number of
PagerDuty incidents that can be resolved in a single cycle, but the process is
subject to PagerDuty API throttling.
Checkpoint and State Management
The PagerDuty Notifier CloudBlade uses a checkpoint system to manage its
state, which affects its behavior when paused or disabled.
- Checkpoint: The CloudBlade saves a checkpoint that marks the last
closed incident it processed, allowing it to resume from that point in the
next cycle.
Pause and Resume: If you pause the CloudBlade, the
checkpoint is retained. When you enable it again, it will resume from
where it left off.
Disabling and Enabling: If you disable the CloudBlade,
the checkpoint is lost. Upon re-enabling, the system will look back at a
20-minute window for closed alarms to resume processing.
Extended Pause: If the CloudBlade is paused for a very
long time, the maximum look-back time is set at 90 days.