cancel
Showing results for 
Search instead for 
Did you mean: 
Data Engineering
Join discussions on data engineering best practices, architectures, and optimization strategies within the Databricks Community. Exchange insights and solutions with fellow data engineers.
cancel
Showing results for 
Search instead for 
Did you mean: 

unity catalog system.access.audit lag

angel_ba
New Contributor II

Hello,

We have unity catalog enabled workspace. To get the completion time of a pipeline that runs multiple times a day, I am  checking system.access.audit table. Comparing the completion time of the pipeline compared to other pipeline time I am creating an alert. But I noticed the audit table has a lag of about 15 minutes i.e. the table dooesn't get updated until after about 15 minutes and as a result a false alert gets raised. How the lag can be handled?

Thanks

Bhawana

3 REPLIES 3

daniel_sahal
Databricks MVP

@angel_ba 
System tables are still in public preview thus there are some limitations, one of them is a blocker for your use case.

Currently no support for real-time monitoring. Data is updated throughout the day. If you don’t see a log for a recent event, check back later.



https://docs.databricks.com/en/administration-guide/system-tables/index.html#known-issues

rabeh_aloui
Visitor

@daniel_sahalThis table can be partitioned by date in the future, it will be easy to track events by day ?

Raman_Unifeye
Contributor III

@angel_ba - This is expected/designed behaviour.

Audit logs are ingested into the system tables asynchronously. Databricks batches these events befor surfacing them in UC system tables. 

Alternate (prhaps) the best way is to use Job API for start/completion time (I presume you are using Jobs for the pipeline). it updates almost instantly.

Second alternate way is DLT event logs itself in the DLT Catalog.schema.event_log


RG #Driving Business Outcomes with Data Intelligence