You Asked

Hi team,

I see spikes in oem for wait event latch undo global data .

This is on insert statement , having concurrency of 50 Inserts in one second

Due to heavily loaded db ash report takes high time

It would be helpful if you share some step
How to trouble shoot this wait event

Basically need some information regarding wait event , what might be possible causes and how to drill it further

Regards,
Krishna

and Connor said...

We store information about the state of undo segments in the SGA. Every time a session wants to know about undo state, it will access this and hence take the latch.

Check out MOS note 1451536.1.

Look at:

Increase the undo tablespace size to make more free space available in the undo tablespace
OR
Reduce the UNDO_RETENTION value, so less undo is being retained

Rating

(1 rating)

Is this answer out of date? If it is, please let us know via a Comment

Comments

Krishnaprasad Yadav, February 19, 2019 - 3:53 pm UTC

Hi ,

I guess commiting transaction would also helpful ,

Can you help me setup case demo for it , if feasible please

February 20, 2019 - 3:36 am UTC

This is not really about committing, although that could have an indirect effect in that some undo may be released earlier.

In terms of resolution, I would piecemeal increase your undo_tablespace size and check for similar reduction of the wait.

More to Explore

Administration

Need more information on Administration? Check out the Administrators guide for the Oracle Database

Breadcrumb

Question and Answer