Sorry, I guess I was reading too quickly also. I saw the 10046 and skimmed the rest. I assumed a session trace and didn't check the syntax. I'll try to be more careful before responding in the future.
Henry -----Original Message----- Sent: Wednesday, January 30, 2002 11:56 AM To: Multiple recipients of list ORACLE-L On Wed, 30 Jan 2002, Babich , Sergey wrote: > However, I made the following entry into initXXXX.ora file: Event=" > 10046 trace name errorstack level 12" and it rebooted over the > weekend. Then I got a call from my boss on Monday morning (I was > home sick) letting me know the system had come to a crawl and was > producing a trace file every minute or so. Eventually, after some > speculations, I told him to rem out that entry and reboot. That > resolved the issue, but left me puzzled. Oh, boy, I couldn't understand who would tell you to do that on a production system, so I looked back in the archives and found this exchange: Sergey writes: > Since I didn't expect that to happen, I had just SQL trace turned on > for that particular session. Do you suggest entering event="10046 > trace name errorstack level 12" into initXXXX.ora? Henry writes: > Couldn't hurt. Also why not do some deltas of your session > statistics (before and after snapshots) I wish I had been paying more attention when this exchange was happening, as I would have said something. It can and does hurt, and you should never set such an event system wide. Furthermore, the event syntax makes no sense. I don't know where you got the syntax for the event, but it is really wrong. What you are essentially asking for is for every process to dump a full errorstack (call stack trace, argument/register address dump, cursor dump, heap dump and PGA dump), every time a process starts up (same as setting 'immediate trace name errorstack...'). Furthermore, this event does not produce a 10046 trace. It just gives you an errorstack. The correct syntax for a 10046 trace is: 10046 trace name context forever, level <level> This should probably not be set instance-wide in the init.ora in a production system. It uses large amounts of resources. Oracle Support has from time to time tried to convince me to set this event instance-wide, but they are mistaken and misguided. It should be set in the session you want to trace, either using 'alter session set events,' 'dbms_system.set_ev(),' or 'oradebug event...' The correct syntax for an errorstack is: [immediate|<errno>] trace name errorstack, level <level> This can be set safely at the instance (init.ora) level if you have it dump on error. If you have it dump on setting the event, as you did, ever server process will have to do this dump when it starts, bringing even a moderately utilized system to its knees. -- Jeremiah Wilton http://www.speakeasy.net/~jwilton -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Jeremiah Wilton INET: [EMAIL PROTECTED] Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051 San Diego, California -- Public Internet access / Mailing Lists -------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Henry Poras INET: [EMAIL PROTECTED] Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051 San Diego, California -- Public Internet access / Mailing Lists -------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
