DBC - CURRENTPERM EXCEEDS MAXPERM

Database
Junior Supporter

DBC - CURRENTPERM EXCEEDS MAXPERM

I had a power outage and of course Teradata was running. 

Teradata - 13.1

I'm running Terdata under VM player and there apparently are no issues with Teradata starting and stopping...

I can logon and run queries e.g. with SQL Assistant but when I attempt to compile 2 procs using BTEQ I'm now getting a msg that there is not more room in database DBC. Interesting in that I'm logging on in BTEQ as a different user. I can compile 1 proc with no problem but more than that... no go. 

I ran a space query and DBC is showing a larger value for currentperm than max perm.. its the only database to show this..

I've also noticed that my response time has slowed considerably since the outage. So, anyone have any ideas how I can fix the DBC issue?

thanks

Tags (1)
2 REPLIES
Teradata Employee

Re: DBC - CURRENTPERM EXCEEDS MAXPERM

Some things charged against DBC in space accounting are not subject to PermSpace limits, other things (like DDL inserting rows in the tables when you create an object) observe the normal rules and fail if they need to acquire more space. In particular, check the size of DBC.TransientJournal (query DBC.TableSizeV). There could have been a lot of work requiring rollback after the crash, and it can take quite a while. If TJ is large, you may want to use rcvmanager utility to see what is holding the space and decide whether you want to cancel the rollback and drop the table. There may be log tables from whch you can purge some data to recover space - ResUsage, DBQL, EventLog, etc. Or if you have taken most of the DBC PermSpace and allocated it to other databases, you may need to move some back.

If the sum of tables CurrentPerm in DBC (from TableSizeV) is much less than the database CurrentPerm (from DiskSpaceV), then running UpdateSpace utility may solve your problem.

Junior Supporter

Re: DBC - CURRENTPERM EXCEEDS MAXPERM

Hi Fred... sorry I took an extensive length of time to respond to you... fighting some gremlins... I reviewed the DBQL tables and they were fairly large, purged most of their contents and currentperm seemed to right itself.. also reviewed the logging rules and made some adjustments.....so, currently good to go.... thanks again for your response...