⚑ Deus Ex Machina ➽ Eventlog Lookup

Deus Ex Machina » Eventlog » Event 514 - ESE

ESE - 514


Description »
Information Store (4788) Storage Group 1: Log sequence numbers for this instance have almost been completely consumed. The current log generation is 918000 (0x000E01F0) which is approaching the maximum log generation of 1048559 (0x000FFFEF), there are 130559 (0x0001FDFF) log generations left to be used. To begin renumbering from generation 1, the instance must be shutdown cleanly and all log files must be deleted. Backups will be invalidated. 
Data formatted as » None

This event will be logged on servers running Exchange 2000/2003 when the Exchange log sequence numbers begin to run out. This is logged in the hope that you will reset the sequence numbers, and avoid an unplanned outage on your Exchange server such as that described in MS KB830408. As explained in The Exchange Team Blog, the first warning will appear when a storage group reaches log sequence number 0xE0000 (917504) - which means you've used just over 87.5% of all available log sequence numbers. For your fourth storage group, the log file would be called E03E0000.log. The event will be logged again for every 1,000 log sequence numbers used on that storage group. When the storage group reaches log sequence number 0xFF000 (1044480), meaning you've used 99.61% of all available log sequence numbers - the entry will be logged once for every 10 log numbers used. Once all numbers are exhausted, the store will dismount, and event ID 1159 from MSExchangeIS will be logged. And you will be sorry.

If you look after Exchange 2000/2003 servers, this eventlog entry is definitely worth trapping for.

Here's a brief change plan I wrote to reset some logs on a clustered Exchange 2003 server recently:

Part 1 - Dismounting the store and databases.

Assuming SG2 (steps will need to be repeated for each Storage Group):

If you see "State: Clean Shutdown" for all databases, it is safe to proceed. Skip to part 3 (resetting the logfiles). If you see "State: Dirty Shutdown" for any database, it must be recovered as describe in part 2...

Part 2 - Recovering databases in 'Dirty shutdown' state:

If you have databases which are not in a 'clean' shutdown state, perform the following steps:

Part 3 - Resetting the Sequence