High Latency on ServiceNow Incident Calls (/api/now/table/incident)

[IbtissemCh]

Hi All,

We are experiencing performance issues with calls to the ServiceNow Table API:

POST /api/now/table/incident

We have noticed that a significant number of requests are taking more than 20 seconds to complete, with some requests reaching 27 seconds or even longer.

Since these calls are part of an asynchronous processing flow, such response times can potentially block or delay downstream consumers.

We are trying to understand whether this behavior could be related to:

• Payload size limitations
• Data volume or table growth
• Recent platform changes or updates
• Any known performance considerations for the Incident Table API
• Is there a recommended payload size limit for POST requests to /api/now/table/incident?

Also, has anyone observed similar latency issues recently or is aware of any ServiceNow-side changes that could impact API response times?

Any guidance would be greatly appreciated.

Best regards,

Ibtissem

[zfrye]

Our users are seeing pretty bad latency on anything that hits an API endpoint. Started with us on Australia and its been a pain to troubleshoot.

[Kieran Anson]

For POST, it's generally best to go down the route of the Import Set API and configuring async transformation to avoid the penalties of a synchronise insert.

That aside, there are a few things that can impact processing time. In Zurich Patch 5 onwards there is a known issue whereby Deny Unless ACLs have a performance impact that is under investigation. This is observed with excessive "slow acl" warning messages in the log. 

Rest debugging would be a starting point for your issue, assuming you can recreate the issue in a non-production environment. Transaction call chain logging is also another option to see what scripts are being invoked. 

If you look at an example transaction record, you'll see a breakdown for SQL time, ACL time, etc. Does one seem particularly weighty?

[yashkamde]

Hello @IbtissemCh ,

As per my understanding likely not a payload size issue, Incident table doesn't have a documented size cap that small/medium payloads would hit. More likely causes:

• Semaphore/thread queueing if requests run in parallel
• Synchronous Business Rules firing on insert. 

Also Refer this :

Troubleshooting inbound integrations performance for ServiceNow administrators and developers 

If my response helped mark as helpful and accept the solution.