The interval has no effect on the number of or the possibility of getting deadlocks. It's an independent process (thread) that pops up on the value (10000ms) notes if there are any deadlocks and resolves them. Changing the value will only resolve them quicker or slower but will not affect whether they happen or not. By the way, a 180 secs. for lock time out *may* be a bit long. If you have an application that "hangs" or does not commit frequently, anybody else requesting a conflicting lock would have to wait 3 mins. before being rolled back. Granted, you hope no appl. will not commit inside of 180 secs. but still, you might want to rethink that one! HTH, Pierre. KG Sivaramakrishnan wrote: > ... > 2) he inerval for checking deadlock is 5000 ms. Currently we have 10000 ms. Still >we get deadlock/timeout. > Will the frequency of deadlock/timeout increase if I change it to 5000 ms. ?
begin:vcard n:Saint-Jacques;Pierre tel;cell:514-233-8679 tel;fax:514-737-1268 tel;work:514-737-4515 x-mozilla-html:FALSE org:SES Consultants Inc. adr:;;233 Simcoe Cr.;Mount-Royal;QC;H3P 1X1;Canada version:2.1 email;internet:[EMAIL PROTECTED] title:President fn:Pierre Saint-Jacques end:vcard
