Re: Library cache lock question

From: Steve Adams <steveadams_at_acslink.net.au>
Date: 1998/01/22
Message-ID: <34c6e713.115395072@newsserver.trl.oz.au>#1/1

Hi Ronald,

This is what is referred to in the documentation as a "breakable parse lock", but if I remember correctly the details are rather scant (as always). These are not locks in the normal sense -- they are what Oracle calls state objects -- that is, a library cache object representing the interest that a session has in some other library cache object, such as a table or user definition, or an SQL statement. These library cache lock state objects are themselves subject to locking by sessions in either exclusive or shared mode. The library cache lock wait event indicates that the target state object is presently locked in an incompatible mode. If it happens again, get the id2 value from v$session_wait (which is the address of the lock state object) and take a library_cache dump at level 4 -- also if feasible, a systemstate dump at level 10 may be helpful.

Hope this helps,
Regards,
Steve Adams
steveadams_at_acslink.net.au

---

On Tue, 20 Jan 1998 23:18:01 -0600, Ronald Boggs <rboggs_at_primary.net> wrote:

>I was executing a large batch of grants through SQL*Plus and it stopped
>in the middle. According to v$session_wait, it was waiting on a
>"library cache lock"
>
>Can someone explain or point me to the doc?
Received on Thu Jan 22 1998 - 00:00:00 CST