Audit logs in Notebooks

[Poojitha_P]

Microsoft Fabric / Spark notebooks, writing small DML operations (insert/update) to an audit log table can be slow on spark cluster taking long time (30–120 sec sometimes) compared to sql db and server.

On optimising this performance would be efficient to capture notebook process status on audit log table.

[v-saisrao-msft]

Hi @Poojitha_P,

Checking in to see if your issue has been resolved. let us know if you still need any assistance.

Thank you.

[v-saisrao-msft]

Hi @Poojitha_P,

Have you had a chance to review the solution we shared earlier? If the issue persists, feel free to reply so we can help further.

Thank you.

[v-saisrao-msft · Accepted solution]

Hi @Poojitha_P,

Thank you @Mauro89, for your insights,

Frequent small DML operations from Spark notebooks, such as audit-log inserts or updates, can result in many small files within Delta tables. This leads to increased task and metadata overhead, which can impact performance. According to Microsoft documentation, enabling Optimize Write helps by performing pre-write compaction, making it useful for tables with frequent small inserts or UPDATE and MERGE operations. Additionally, Auto-compaction can automatically compact fragmented files after writes, which helps maintain table performance when there are frequent small writes.

Tune File Size - Microsoft Fabric | Microsoft Learn

Cross-Workload Table Maintenance and Optimization - Microsoft Fabric | Microsoft Learn

Thank you.

[Mauro89]

Hi @Poojitha_P, 

welcome to the community and thanks for your post.

Iam not sure about the question you have, may I ask you to clear thinks up a bit?
If it is more of a feature request I recommend you to post this in the ideas section of this community here: Fabric Ideas - Microsoft Fabric Community

Best regards!

PS: If you find this post helpful consider leaving kudos or mark it as solution