Importing to Copy2 when Copy1 is unavailable

[Christian_Jones]

We're in OnBase version 16.0.0.20.  We have a UAT disk group with documents already imported configured for Copy1 (Mass) uncommitted, and Copy2 (Mass).  We thought that if network access to Copy1 was unavailable that OnBase would then be able to Import to Copy2. Document Retrieval is working as expected.

Attempting to import via Thick Client = Platter Management error, problem locating the ID file for Disk Group: The uncommitted copy of the Disk Group is currently offline; Error: GetIdFile 2: -3

Attempting to import via Web Client = The platter for this document is currently unavailable.

In testing a Disaster Recovery situation, we took Copy1 offline.  When attempting to retrieve a doc from the Disk Group, it is successful and shows Status=NO for Copy1 and Status=OK for Copy2 within Properties | Locations for said document.

Under what condition(s) would we be able to Import a doc to Copy 2 if Copy 1 is not available?

Or, asked differently, what are the "rules" for OnBase to Import to Copy 2 if the Copy 1 location is not available (for various reasons, i.e. no network rights to Copy 1, or that "drive" is technically not reachable, etc.)?

...

Going further, if there is a failover OnBase environment... Copy1 is site A and Copy2 is site B. Site A and B have their own independent fileshare UNC naming convention say site A is Chicago and site B is NYC.

Should site A go down, how would we fail over to site B without having to rename all of the thousands of disk group platter paths --- we have an Alias to use for site A and B, but it sounds like there is still a manual change at the DNS level - or does anyone know of an automatic method at the DNS level ... ?

...

Copy 1: \\site.A.Chicago.primary01.failover.com\uat$

Copy 2: \\site.B.NYC..secondary01.failover.com\uat$

In other words,

Can the DNS be setup to go to the normal server mapping when that server is available, but go to the failover server when the normal server is not available?  This would allow for an automated failover without the DNS change being a manual intervention.  If this is not possible, we're curious how other failovers are handled in DNS - if anyone has experience with this that is. 🙂

Reference(s):
https://www.onbase.com/community/technical_communities/platter_management/f/platter-management/10486...

[Michael_Reindel]

Hi Christian,

We have recently released a High Availability/Disaster Recovery document here: https://www.onbase.com/community/technical_communities/virtualization/m/resources/20192

When OnBase Copy 1 is not available, documents cannot be ingested into the system. Promoting Copy 2 to Copy 1 is an option, but requires modification of the ID files, validation of the data on Copy 2, and the creation of another Copy 2 so that a second copy of the data exists while in DR mode.

If you experience a storage failure so that Copy 1 is unavailable, I would recommend creating a new location to house Copy 1 in the primary datacenter. This method ensures that committed data ingested before the failure is retrievable from Copy 2, and new data can be written to the new, temporary Copy 1 and replicated to Copy 2 during commit.

In any case, if you experience any loss of this type, you should contact your first line of support for the best recommendation for the situation.

Mike

[Christian_Jones]

Hi Mike,

Thank you for the response. I am reviewing it and the documentation you've pointed to.