Upgrading from v7.20

When upgrading from v7.20, you will NOT need to restore the alarm data files (ALARMSAV.DAT and ALRMSAVEINDEX.DAT) under most circumstances. Citect SCADA 2018 is equipped to read this information from the redundant v7.20 (SP5A or greater) server that is still not upgraded.

To upgrade from v7:20:

1. Add the following parameter on the .INI file to all your server nodes before you start the online upgrade.

[LAN]EarliestLegacyVersion = 7200.

Restart the servers after adding the parameter for the changes to take effect.

2. Shutdown SCADA runtime on the primary server.
3. Upgrade Citect on this server according to the offline upgrade procedure.
4. Restart the primary server. It is now upgraded.
5. Now, the Citect SCADA 2018 server will build the new alarm database, and will import the historic data from the Standby v7.20 server.
6. Check the status of the alarm server synchronization using the Alarm Server Kernel, on the Main Window:

• W hen the Alarm Servers synchronization starts you should see the following message:
  Alarm: Peer update request sent.
• Then you should see a number of messages with Update packets (number is dependent on your Alarm historic events and configuration).
  Alarm: Update packet XXXX received.
• Finally, the following messages will indicate that the synchronization has been finalized successfully:
  Alarm: Database objects state synchronization completed.
  Alarm: Database is initialized, preparing to Start the Alarm Engine.
  Alarm: Starting Alarm Engine
  Alarm: Server startup complete.

7. If you find that your Alarm Server synchronization is not completing successfully, place the ALARMSAV.DAT and ALRMSAVEINDEX.DAT on the [Alarm]SavePrimary directory.
8. Upgrade your client nodes one by one.
9. Once you are confident that synchronization of alarms, trends etc., is complete, and that your v2018 clients are working correctly, shutdown runtime on the Standby server.
10. Upgrade Citect SCADA on this server according to the offline upgrade procedure.
11. Restart the standby server. It is now upgraded.
12. Once the standby server is running fine, check for hardware alarms when it is connected to the primary server.
13. Check functionality of the system as a whole.
14. Finally, test redundancy by switching off the primary server and checking that the standby server takes over and clients switch over.

Special Considerations

Custom Alarm Filtering

The AlarmSetQuery Cicode function was deprecated in v7.30. This means that if you are using custom alarm filtering code, you will most likely need to convert it. Click here for more information about this process.

Historical Alarm Events

Set the [Alarm.<Cluster Name>.<Server Name>]ArchiveAfter .INI parameter to a date prior to the earliest historical event date from which you want to migrate.

Alarm server synchronization during online upgrade

In the event that there is a disconnection or timeout during synchronization between the v2018 and v7.20 alarm servers, follow these steps:

1. Shutdown your 2018 server.
2. Delete the alarm database and re-start it.
3. Wait for the synchronization between servers to finish.

Also, you can increase the timeout using the [Alarm]StartTimeout .INI parameter. This will allow the v2018 server to wait for connection from the v7.20 server.

If you find that the synchronization between the two servers is experiencing interruptions, delete the alarm database, and place your ALARMSAV.DAT and ALARMSAVINDEX.DAT in the [Alarm]SavePrimary directory and the v2018 server will convert the data. However, we recommend always trying the peer synchronization first.

Changes during the upgrade process

Because of the differences between Citect SCADA2018 and v7.20, any actions that happen during the online upgrade process are subject to incompatibilities that are not reconcilable between versions. However, the scenarios are quite particular and should not have a great impact if any, on your SCADA system. Here is a list of such scenarios:

• UserLocation field: In Citect SCADA2018, a record of the UserLocation, that is the IP address, for alarm operations such as acknowledge is available. If an acknowledge occurs on the v7.20 server during the upgrade, the v2018 server will be unable to record the UserLocation, which will be displayed as "0.0.0.0".
• Summary Comments during the upgrade: Comments that you add to an alarm summary record on the v7.20 server during the online upgrade will not be available in the upgraded version.

Published June 2018