Delete lost messages

[Dvir]

Hi There,
During a failed attempt to upgrade the hardware of the production server I have lost the HSM drive to a nasty crush.
It is not a disaster to the company, but I have to tidy up things in order to let the users carry on working without errors.
Currently if they go back in their inbox previous to the HSM cutover date, they can see messages which they cannot open, since the BLOB of the message is lost.
This is obvious however, I tried few mailboxes and done the reindex mailbox, however it didn't solve the problem. It still doesn't delete these messages or pointers to the messages.

Does anyone have an idea of how to remove all the "lost" messages from all the mailboxes in one process?

Thank you very much for your help,
Best Regards,
Dvir

[uxbod]

Well, first of all to ascertain the full impact you could run

su - zimbra
zmblobchk

It would also be helpful if you would update your profile with the following output so we know which version of ZCS you are running

su - zimbra
zmcontrol -v

Well, there is probably a easier way of doing what I am going to describe but here goes anyway

su - zimbra
zmvolume -l    * this will list all your volumes so you can see which is your HSM
mysql
use mboxgroup1;
select * from mail_item where volume_id = '3'   * obviously change the volume_id to be your HSM one.

You could then delete all the mail items that are in your HSM but do not have any data files. You will need to do this for all the 10 mboxgroups and then re-index once completed. Alternatively, you may just be able to remove the HSM store and all the referencial data will be updated accordingly.

NO warranty comes with any of this as I personally keep good backups

[Dvir]

Using the latest NE Version.

5.0.16

On RHEL 4 Machine.

Will the blob check also remove or only display the problem?
Is there a flag to activating a removal of invalid blobs?

[Dvir]

Originally Posted by uxbod

NO warranty comes with any of this as I personally keep good backups

You right, me too, it's all backed up on the users' machines and already done archiving on these machines. That's why I wrote that it is not really a problem that the information is lost.
I didn't expect the storage device to fail on me like that, but nothing really important is lost, as all the important mailboxes are backuped seperately.

Thank you for the input, will give it a shot on Monday.