[quote=Marco Kalter][quote=Kirurgs]BUT for some strange reason now plsqldev.exe process eats whole cpu all the time after connection is made, cpu usage varies from 65% to 99%.[/quote]

Does the CPU usage go down eventually?

Can you test what happens if you go to Tools > Preferences > Code Assistant and disable the "Describe Context" option? Does this clear the CPU usage after a restart? [/quote]

Hi!

I'm on vacation so I check this very rare.
CPU won't go down, it starts even when there are no windows opened at all, just plain PLSQLDev. I went to check sessions and no query is running.
That check box didn't help either.

Hi!

v9.0.1 (1613) have the same CPU usage bug. I still see no queries being performed while CPU usage is high.
DB ver. is 11gR2.

Edit: In 9.0.1 CPU went down after under 5 mins (approx 3-4). Do You have clue why is that?

I'm not sure why this would happen and we cannot reproduce this.

Does the CPU usage go up again when you connect to a different database after the CPU usage goes down?

Hi!

I think I understood where the problem is
See, we are working with Oracle eBusiness Suite which is really large product of Oracle, I took statements from CANames and executed them with the schema I logged in (APPS), results are: objects (1st stmnt) ~ 80000, synonyms (2nd stmnt) ~ 50000, links (3rd stmnt) = 7.
It seems that PLSQLDev tries cache them all and do smth with them and uses a lot of CPU then. I added 1=2 to the statements and no excessive CPU usage anymore but code assistant didn't work.
It seems that in v8 code assistant was querying stuff on-demand but now it tries to cache it...
Is there a setting which reverts to old behavior or setting which prevents caching all of the stuff at once?

regards
Kirurgs

We'll check it out. Maybe there is something we can do to improve this.

Can you try this version and let me know if the CPU usage is reduced?

Hi!

I have tried it and it is a lot better now.
However, when I'm trying to use code assistant on tables which are used via synonyms they don't appear to be visible for some time, after couple of mins after logging in they are, but not right away as those who are in the current schema.

Personally me, I can live with that, but if You can improve this, it would be nice.

My guess would be that querying the synonyms is slow. You can verify this be running or modifying the CANames.sql file in the PL/SQL Developer directory.