[noname]

Sharing this known issue - DT511330.

Overview

Listed in 5.0.3 Release Notes:
In an environment that included an IBM® Lotus® Domino® cluster and users with a large number of unread messages, when the cluster failed over, if the Note IDs for the replicated messages were not the same, the BlackBerry Enterprise Server reprocessed all unread messages and it might have taken a long time before devices received the messages. (DT 511330)

There is also an IBM Tech Note "Potential hang while upgrading the BlackBerry Enterprise Server on IBM Lotus Domino to 5.0.3" where IBM logged SPR JPAI8E4NTW:
https://www-304.ibm.com/support/docv...id=swg21470828

Note: This appears to be an Domino issue as the Tech Note states:
Fix: The fix for SPR JPAI8E4NTW is in 8.5.2FP2 and 8.5.3. The mail servers which interact with the BES must be upgraded. The BES server does not need to be upgraded.


Details

What happen is that during a mail cluster server failover, if the mail documents have different NoteIDs between the mail file replicas, then the BES could possibly see a different status for a given message in the current replication than the previous replication. This could cause the status change to be reconciled to the BlackBerry smartphone.

Also because of this issue, the BES could potentially have a lot of read/unread marks reconcilations and as a result cause high CPU on BES Domino server; given that there are BES users with large mail files.

To identify this issue, you can see that there is multiple reconcilation for a specific message.

[MAGT]
[30613] (01/01 10:46:33.834):{0x1B28} {User Name/Domain} Mail Server CN=Mail1/OU=SRV/O=Domain failed over to CN=Mail3/OU=SRV/O=Domain
[40395] (01/01 10:47:33.615):{0x1B28} {User Name/Domain} OTAFM (RID=-1234567890): message status updated to <unread>, constructing notification for the device
...
...
[30613] (01/01 11:07:55.530):{0x1AEC} {User Name/Domain} Mail Server CN=Mail3/OU=SRV/O=Domain failed over to CN=Mail1/OU=SRV/O=Domain?
[40395] (01/01 11:16:34.533):{0x1B34} {User Name/Domain} OTAFM (RID=-1234567890): message status updated to <unread>, constructing notification for the device
...
...
[30613] (01/01 11:29:59.366):{0x1B10} {User Name/Domain} Mail Server CN=Mail1/OU=SRV/O=Domain failed over to CN=Mail3/OU=SRV/O=Domain
[40395] (01/01 11:30:52.570):{0x1B10} {User Name/Domain} OTAFM (RID=-1234567890): message status updated to <unread>, constructing notification for the device
...
...
[30613] (01/01 11:34:18.587):{0x1B14} {User Name/Domain} Mail Server CN=Mail3/OU=SRV/O=Domain failed over to CN=Mail1/OU=SRV/O=Domain
[40395] (01/01 11:49:08.998):{0x1B18} {User Name/Domain} OTAFM (RID=-1234567890): message status updated to <unread>, constructing notification for the device

Source: KB21083-Read or unread status changes when failover occurs

[noname]

Just noticed that there is incomplete snip.

Listed in 5.0.3 Release Notes:
In an environment that included an IBM® Lotus® Domino® cluster and users with a large number of unread messages, when the cluster failed over, if the Note IDs for the replicated messages were not the same, the BlackBerry Enterprise Server reprocessed all unread messages and it might have taken a long time before devices received the messages. (DT 511330)

To resolve this issue, verify that you are using Lotus Domino 8.5.2 or later.

Note: RIM technical writer is not stating it clearly for the upgrade. It should read as:

To resolve this issue, verify that you are using Lotus Domino Mail Servers 8.5.2 or later.