Seeing Lots Of Repair Requests After Upgrading From MS 7u4 To MS 7u5 (Doc ID 2079614.1)

Last updated on MAY 03, 2021

Applies to:

Symptoms

Seeing a major performance impact on the server after upgrading from MS 7u4 to 7u5

Even after shutting down the dispatcher, enabling our customized panic settings (less logging, no snapshots, etc.), configuring msprobe to run with -n, and then upgrading.  We typically have a period of 1-3 hours where imapd threads are all totally maxed, banner and fetch are slow (upwards of 5 minutes for each phase at times) but the store does continue to function, just BARELY (and if msprobe did not have -n it would be restarting it, making things worse).  During this period, even with dispatcher down and no new mail coming in, we often see hundreds to sometimes thousands of repair requests building up in the maintenance queue.  Then it gets over the hump (perhaps after finishing upgrading the bulk of the active users) and everything becomes fine again.

Changes

Upgrading from MS 7u4 to 7u5

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.