Video Screencast Help

watchfile talbe not clearing - based on archive bit

Created: 09 Jul 2013 | 14 comments
Sarah.Seftel's picture

EV version 9.0.3, with the watchfile table fixed applied.

partitions set to check for archive attribute, but the watchfile table keeps growing until we reset the admin service.

what else can be checked here?

what is the trigger for the watchfile table to clear the rows once partitions set to archive bit attribute?

 

Thanks,

Sarah 

Operating Systems:

Comments 14 CommentsJump to latest comment

JesusWept3's picture

going in and out of backup mode will cause the scan, you can also set the scan times in the properties of the Vault Store Partition itself, and there are reg keys that can be used to control how often it checks for backed up items.

Unfortunately its just not practical to have EV Constantly scanning each and every item to see if the archive bit has been removed

Sarah.Seftel's picture

Hi

System is getting in and out of backup mode with the EV agent that BackupExec have.

But still - it does not clear the table.

EV_Ajay's picture

Hi Sarah,

Could you let us know the Storage Type : NTFS / EMC Centera ?

 

In the Partition Properties "Backup" Tab have the setting "Scan partition every : 60 minutes". If you reduce this value it will decrease the next scan time. This will check the value in the Watchfile Table for ItemSecured = 0. ( It means that item is backup). Then it will remove the entry from the Watchfile table.

 

 

 

Thanks,

Ajay

Sarah.Seftel's picture

Hi Ajay,

The storage type is NTFS.

The partition is set on "Use archive attribute" so the scan for trigger file is not set.

Advisor's picture

Did you confirm if the BackupExec agent is actually resetting the Archive Bits? You can use attrib command to check the "A" attribute of dvs files on the partition which was backed up. If you see "A" attribute on the file, then its backup which is not working as expected. If you see that archive bit has been reset by backup application but still entries are there in watchfile table, then confirm the table entry column backupcomplete or itemsecured kind of. If its zero then its not updated by Storage File Watch process, which should be trigered after clearing the backup mode.

DTrace of Storage File Watch will give you insight of whats happening with the process.

Sarah.Seftel's picture

Backup clears the archive attribute.

watchfile table shows itemsecure=1, it only cleares the table after I reset the services.

I will trace it.

Advisor's picture

Ah, ok. So it looks like, StorageFileWatch does not trigger after clearing the backup mode. When you restart the storage service, it triggers the file watch and then clears the entries from the table.

Get a dtrace of this process while clear the backup mode, along with StorageManagement process.

Sarah.Seftel's picture

I think I should schedule a dtrace for the time system is getting out of backup mode, as it should trigger the file watch and it's not doing it.

What will I get by tracing the system while I'm resetting storage service? as I already know that reset of storage service fixes the problem...

The issue is that the trigger is not working on regular bases...

Am I correct?

Sarah.

Advisor's picture

Yes thats correct. If trace the storagefilewatch process while resetting the storage service, you will get the working log which then can be compared with the log generated during backup mode clear operation.

By default, when you restart the storage service, it triggers the file watch process. I guess thats not happening after backup mode clear operation.

EV_Ajay's picture

Hi Sarah,

Thanks for your reply.

  • Use the archive attribute. Files on the partition whose archive attributes are cleared are considered backed up and Enterprise Vault removes the corresponding safety copies. This option is appropriate only if your backup software clears archive attributes after backup.

 

It means need to check whether your backup software is clearing the "A" attribute after completion of the backup. Make sure you are running the Full backup of the Vault Store Partition.

To trigger the Storage File Watch process to check the item backed up or not , need to restart the Enterprise Vault Storage Service. After that it check for the item is secured or not.

Or you can create the batch file to restart the Enterprise Vault Storage Service.

Thanks,

Ajay

Sarah.Seftel's picture

Hi Ajay,

The backup software is clearing up the archive bit attribute correctly.

The EV seems to ignore it. when I restart admin service - it clears the table.

EV_Ajay's picture

Hi Sarah,

Could you try to reinstall the Backup Exec Agent on the Enterprise Vault Server and then test the same thing.

 

Thanks,

Ajay

GCSPT's picture

is this the same system you had the Journal Delete records not clearing down.

Do you get the 7080 event with no errors coming out of backup mode. Then a 7102 storagefilewatch event commenting on the scan being stopped.

What are the DB related warnings in the VAC>directory>status console.

Symantec EV Support Engineer EMEA

EV_Ajay's picture

Hi Sarah,

It seems like you backup software clear archive bit correctly but Backup agent is not able to send trigger for scanning to ev server. Hence i suggest to reinstall the backup exec agent on ev sever and test it.

 

Thanks,

Ajay