You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@tomcat.apache.org by Nikolaos Giannopoulos <ni...@solmar.ca> on 2003/06/03 18:25:11 UTC

SOLVED: Determining Solaris OS JVM stability

Steps:
======
(1) Solaris OS *upgraded* from 2.6 to 8.
(2) OBSERVED: Java 1.4.1_01 processes simply *disappear*
(3) Java (cluster) patch installed
(4) OBSERVED: Java 1.4.1_01 processes *STILL* simply *disappear*

Steps:
======
(1) Solaris 8 *clean* installed.
(2) OBSERVED: Java 1.4.1_01 processes simply *disappear*
(3) Java (cluster) patch installed
(4) PROBLEM SOLVED

Scope:
======
Problem affects any Java process - Tomcat, application, etc... and may not
exhibit itself unless the application uses a specific degree of
multi-threading.

Wild Guess:
===========
This leads me to believe that the Java patch works well on a system with
Solaris 8 cleanly installed on it BUT not on a system that was upgraded from
Solaris 2.6 to 8.  It could be - I'm guessing - that in the upgrade process
an older (threading dependent or other) library (from 2.6) is not migrated
(to version 8) as the Java patch may expect.

In addition, I wonder how much testing is done with the Java patch on a box
that is upgraded from Solaris 2.6 to 8 - I imagine none; as one expects that
once a box is upgraded to 8 that it is functionally equivalent to a clean
installed 8 AND therefore tests would focus on clean installed 8 boxes.  All
in all this makes sense but makes problems of this nature difficult to solve
and track down.

Luckily we had 5 boxes that run our software.  Un-luckily this cost us about
6 weeks of total time (on two separate incidents) to track down.

Hope this helps someone down the road.

--Nikolaos



---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-user-help@jakarta.apache.org