How to migrate or move the historian server to a new server

This GE Article describes the process CLICK HERE.

10,000 foot view of the process

At a high-level, the key is that the .ihc file is the cornerstone.

The .ihc file contains all the configuration, the list of archives, tags, collectors, etc.

When your old server is called “SERVERA” and new server is called "SERVERB” your old .ihc file will be named SERVERA_Config.ihc

If you can copy everything from SERVERA to the SERVERB folder (.IHA files and .IHC file) you rename the .ihc file to SERVERB_Config.ihc

And when you start the data archiver service, it will read the ihc file that matches the server’s hostname.

More verbose detail about the process

The long winded process involves more detail about the following:

You start with a clean operating system on the new machine (Don’t clone it and upgrade it)

When you run the historian installation on the new machine, it will ask for the location of the “proficy historian data” folder.

If your old server had d:\proficy historian data, your new server should be d:\proficy historian data. And we assume you would already have that drive created with enough disk space for your existing archives to be copied there twice.  This allows to have (1) a folder that will only be used as a repository for the old archives and never to be read or written to and (2) a backup folder that the existing archives (which will be upgraded) and any new archives will be stored.

Make a note of the paths of every datastore. The .ihc file expects the datastores to be in the same directory that they were in on the original machine.

Notes on copying files:

  • Get a catalog of the current archives on the old system. You can use excel addin or ihSQL.exe. For ihSQL.exe, you can use the following select statement and copy/paste into notepad and bring it into excel and transform “text to columns”

    select * from iharchives
  • You cannot copy the .IHC and .IHA files from the source computer while the data archiver is running. If you can afford some downtime (the collectors will buffer) you could stop the data archiver service and start copying. To save time, only copy the files in the catalog from the ihArchives table. The others can be placed in a backup folder because they are not online archives.

  • If you cannot afford downtime, you can use the ihArchiveBackup.exe to back up everything. The first thing you want to do is to make sure your backup directory has been set to a different folder than the archives folder.

1.) Click Datastores

2.) Click Datastore Options tab

3.) Click Default Backup Path and set it to a DIFFERENT folder than the default archive folder.

  • If you skip the step above, backing up the .IHC file will go into the archive folder and you STILL won’t be able to copy it out.

  • The iharchivebackup.exe with the -c option will make a backup of the .IHC file. To back up the archives, I usually run one command per archive and put it in a batch file that looks like the following

"c:\Program Files (x86)\Proficy\Proficy Historian\Server\iharchivebackup.exe"-a WCHMDCSHIST_Archive991 "c:\Program Files (x86)\Proficy\Proficy Historian\Server\iharchivebackup.exe"-a WCHMDCSHIST_Archive992 "c:\Program Files (x86)\Proficy\Proficy Historian\Server\iharchivebackup.exe"-a WCHMDCSHIST_Archive993 "c:\Program Files (x86)\Proficy\Proficy Historian\Server\iharchivebackup.exe"-a WCHMDCSHIST_Archive994 "c:\Program Files (x86)\Proficy\Proficy Historian\Server\iharchivebackup.exe"-a WCHMDCSHIST_Archive995

Do the move by following the GE article (same article as at the top)

This GE Article describes the process CLICK HERE.

Ready to start the new archiver service?

Before starting the archive service on the new historian, you need to install a license that is compatible with the amount of tags and datastores that you currently have.

Once you have the archives where they belong (in the same path as the original server) and the IHC copied over, and you have a good enough license, you can rename the .IHC file so it has the hostname of the new server. Then you can start the historian data archiver. If it doesn’t stay running, you need to review the log files.

After you get the new server running, you may be experimenting with it for a little while and brainstorming how to “cut over” the collectors.

If your collectors use a hosts file, you could modify the hosts file to change the IP address of the historian.

If not, there is a registry setting that you can change to rename the Historian Server hostname that it is pointing to.

You may have a small gap of data unless you have a strategy to “cut over” all collectors at the same time.

You can only do a cutover “at the same time” if you keep the same historian server name on the new system which can be done with two different subnets and you’d shutdown the old server, then re-ip the new server to the old IP and then all the collectors would connect.

The specific details of the above vary depending on how the old server was set up.

Wow, that is a lot and I think I might want some help!

If you feel better to have expert backup, one of our AutomaTech professional services experts, can help you:

A) On site visit to do the migration for you.

B) Helping with remote access or online meetings to prep and migrate the data to the new server, then do cutover on site.

C) You could also buy a bank of hours, do most of the work yourself and call for help as needed.

AutomaTech Inc.