Checksum errors when performing a failover from Standby Server to Background Server.

book

Article ID: 36499

calendar_today

Updated On:

Products

SUPPORT AUTOMATION- SERVER CA Service Desk Manager - Unified Self Service KNOWLEDGE TOOLS CA Service Management - Asset Portfolio Management CA Service Management - Service Desk Manager

Issue/Introduction

Summary:

In ServiceDesk versions R14.1 or R12.9 - Advance Availability mode, when performing a failover from the Standby Server to the Background server or vice-e-versa, checksum errors appear in the logs that would not allow the ServiceDesk Services to start on the Standby Server or Application Servers.

Due to the checksum errors, data from the custom tables may get truncated and the CA Service Desk Manager Services on the Standby Server or any of the Application Servers may fail to start. This is generally noticed after schema changes are performed using Web Screen Painter (WSP).

The below checksum errors appear in the stdlog:

11/26 16:54:13.18 SBYSERVER pdm_rfbroker_nxd 2868 ERROR ServerStatusMonitor. 1956 Unable to register server BGSERVER (32143) as checksum count 3 would exceed maximum of 2 (7 servers registered) 

11/26 16:54:13.18 SBYSERVER pdm_rfbroker_nxd 2868 SIGNIFICANT ServerStatusMonitor. 1981 (checksum #1) STANDBY server BGSERVER (32143) has checksums ddict(2590889872) Majic(3632789897) wsp.mods(3958148182) 

11/26 19:05:47.81 SBYSERVER pdm_rfbroker_nxd 2868 ERROR ServerStatusMonitor. 1956 Unable to register server APPSERVER4 (32147) as checksum count 3 would exceed maximum of 2 (7 servers registered) 

11/26 19:05:47.81 SBYSERVER pdm_rfbroker_nxd 2868 SIGNIFICANT ServerStatusMonitor. 1981 (checksum #1) BACKGROUND server SBYSERVER (32144) has checksums ddict(2590889872) Majic(3632789897) wsp.mods(4050571555) 

11/26 19:05:47.81 SBYSERVER pdm_rfbroker_nxd 2868 SIGNIFICANT ServerStatusMonitor. 1981 (checksum #2) APPLICATION server APPSERVER5 (32145) has checksums ddict(2075879600) Majic(3632789897) wsp.mods(4050571555) 

11/26 19:05:47.81 SBYSERVER pdm_rfbroker_nxd 2868 SIGNIFICANT ServerStatusMonitor. 1981 (checksum #3) APPLICATION server APPSERVER4 (32147) has checksums ddict(2590889872) Majic(3632789897) wsp.mods(3958148182)  

 

We consider 7 ServiceDesk Servers in Advance Availability (AA) Mode - 1 BG Server, 1 Standby Server and 5 Application Servers.

  • A failover is being performed from the Standby Server (SBYSERVER) to the Background Server (BGSERVER)
  • The current Background Server hostname is SBYSERVER.
  • The current Standby Server hostname is BGSERVER.
  • The Services fail to start on the Standby Server (BGSERVER) and the Application Server APPSERVER4.   

 

Instructions: 

From the above stdlog messages we can see that there are 4 different sets of checksums:

1. (checksum #1) BACKGROUND server SBYSERVER (32144) has checksums ddict(2590889872) Majic(3632789897) wsp.mods(4050571555) 

2. (checksum #2) APPLICATION server APPSERVER5 (32145) has checksums ddict(2075879600) Majic(3632789897) wsp.mods(4050571555) 

3. (checksum #3) APPLICATION server APPSERVER4 (32147) has checksums ddict(2590889872) Majic(3632789897) wsp.mods(3958148182) 

4. (checksum #1) STANDBY server BGSERVER (32143) has checksums ddict(2590889872) Majic(3632789897) wsp.mods(3958148182) 

 

As per design, Service Desk Advance Availability setup does not allow 3 different sets of checksums. Due to this:

  • The Standby Server will not start as it has different checksum than the BG and APP Servers. 
  • The APPLICATION server APPSERVER4 will not start because it has different checksum than BG and other APP servers.
  • So if any of the other APP servers go down, then the services on these APP Servers will not start.

 

Perform the below steps on the problematic Application Server. In this instance on APPSERVER4.

1.Set @NX_RFBROKER_RELAX_CHECKSUM_RESTRICTION=3 in the ServiceDesk NX.env file.

2.Start the Service Desk services on APPSERVER4. 

3.Wait for version control to synchronize files. 

4.Stop services on APPSERVER4. 

5.Remove the set variable @NX_RFBROKER_RELAX_CHECKSUM_RESTRICTION=3 from NX.env 

6.Start the Service Desk services on APPSERVER4. 

7.Check the stdlogs of the Background Server and see if the checksum of this APP server "APPSERVER4" matches that of BG server. 

8.      If this works then perform the same steps on the Standby Server.

 

Note If the above steps do not work, then all the Servers must by physically restarted so that the checksums on all the Servers are matched.

Environment

Release: SDMU0M99000-14.1-Service Desk Manager-Full License
Component: