smon deadlock on TX enqueue - how long to resolve ?

From: GG <grzegorzof_at_interia.pl>
Date: Tue, 20 Sep 2011 18:04:27 +0200
Message-ID: <4E78B98B.6090004_at_interia.pl>

Hi,
 I think I came across

    Bug 7529174 - Deadlock / hang between SMON and foreground process     (Doc ID 7529174.8)

on my 4 node 10.2.0.3 rac .
There was user$ and obj$ tables involved (smon had TX enqueus on them) . I did some killing :) on sessions waitng on row cach id=7 (dc_users IIRC). Well, there was like 7000 ctime on that lock for smon sid . I'm wondering why standard deadlock resolving routine did not worked for that .
So my question is how long such deadlock would lasts (on smon)? Additionaly I saw a lot of session waiting on SS enq (sort segment? ts#=3 = temp) .
Not sure if that was because of smon hang (probably yes but please confirm). So far instance seems stable till next alter user race condition :). Regards
GregG

---

Skoda Octavia z 2009 roku juz za 11 tys. zl http://linkint.pl/f2a36
--

http://www.freelists.org/webpage/oracle-l Received on Tue Sep 20 2011 - 11:04:27 CDT