Scratch all that.. I have figured it out..

The date modified timestamp will generally get updated on an Exchange database when either of the things happen
1. The EDB file size is extended in order to accommodate data that does not fit into whitespace that currently exists in the database.
2. Or the database is dismounted and all open handles to the files are released.

We had a lot of whitespace which meant that the database file did not get marked as modified by Windows.. simple..

Thanks again for the original post..

I have no idea how I have not noticed after all these years that the ebd files are updated at wildly varying intervals. It’s only now, during/after this migration when I was paying close attention to the database file that I noticed that it did not update.

I always understood the write-forward nature of Exchange to be; continuously write to the transaction logs, periodically (like minutes, not days or weeks) write that data into the database and prune the logs when the database has been backed up.

However, I now realize that this was wrong and that the logs can be pruned and the database does not get updated. Soo, where exactly is the updated info held? For example I just referenced a server that we’ve had in production for years. It’s database last updated weeks ago but the transaction logs are removed nightly with the backup job. Where is the data held before it’s written to the database if the logs are removed? Memory buffer, e.g. RAM? If so, what happens if the server suddenly loses power?

Thanks a bunch for the post, it has absolutely helped us; however, after the migration we are seeing a very strange behavior.

The edb file located in it’s new location is not growing. The size has not changed since the migration was complete and the modification time shows the time just after the migration. Mail is in fact flowing in both directions and the transaction logs are growing in number as expected. The transaction logs even prune properly after we run a backup. But still the edb file has not changed.

The exact command that we used is:

Move-DatabasePath “Mailbox Database 1602191925” -EdbFilePath “F:\ExchangeData\Mailbox Database 1602191925\Mailbox Database 1602191925.edb” -LogFolderPath “E:\ExchangeData\TransactionLogs\Mailbox Database 1602191925”

I have been running the following command:

Get-MailboxDatabase -Status | select Name, DatabaseSize,AvailableNewMailboxSpace

The output being:

Name Database Size AvailableNewMailboxSpace
Mailbox Database 1602191925 228.2 GB (245,081,571,328 bytes) xx.xx MB

The ‘AvailableNewMailboxSpace’ value has changed each time, sometime up and sometimes down, but only 10MB this way or that..

This is a production environment with large amounts of email passing through at all times.

Any insight would be very very appreciated..

I don’t have circular logging on so not sure if that would help at this point or not.

I need to move all databases from the passive database server to a new partition because it’s almost full and it’s made out of spanned disks.
Will something like this work?
-stop the replication
-put the server in maintenance mode
-move everything to the new partition
-change the letter to the new partition to be the same as the old one so the edb file path will stay the same – logs are on a different partition
-take the server out of maintenance and resume the replication

I have Exchange 2013 environment with CU19, Please let me know if I can create more than one database on Single Drive? is this supported by Microsoft?

Thanks,
Azad

Those steps are not correct, even with your second comment. I don’t recommend using this method to move databases in a DAG. Aside from the errors, you’ve reduced the redundancy of the DAG by removing a database copy, and you also cause an outage.

Better to create a new database in the new path, with database copies, and move mailboxes to the new database using move requests.

Sorry, forgot one Thing between 2) and 3):

On PASSIVE NODE:
Remove-MailboxDatabaseCopy -identity “Mailbox Database 1\PASSIVE_NODE_NAME”

Dear Paul,
can you add the steps while using a DAG with DataBaseCopies?

1) Suspend-MailboxDatabaseCopy -identity “Mailbox Database 1\PASSIVE_NODE_NAME”

2) robocopy E:\Mailbox Database 1\ “F:\Mailbox Database 1\Mailbox Database 1.edb” /MOVE /SEC

3) Move-DatabasePath “Mailbox Database 1” -EdbFilePath “E:\Mailbox Database 1\Mailbox Database 1.edb” -LogFolderPath “F:\Mailbox Database 1”

4) Mount-Database -Identity “Mailbox Database 1”

5) Add-MailboxDatabaseCopy -Identity “Mailbox Database 1” -MailboxServer “PASSIVE_NODE_NAME”

6) Resume-MailboxDatabaseCopy -Identity “Mailbox Database 1\PASSIVE_NODE_NAME”

7) Update-MailboxDatabaseCopy -Identity “Mailbox Database 1\PASSIVE_NODE_NAME” -CatalogOnly

😉
Cheers
Kory