Views:

Summary



This procedure specifically describes a method to perform a DPX data base consistency check. This DB consistency check is recommended any time the DPX data base is moved to a new master server. This includes moving the data base to a new machine with the same OS or a different OS. Please note: The 'syncui' utility should only be used under the direction of technical support; In this case, specifically to validate a DPX data base migration to a new platform.

Resolution



You will need to invoke a command line on your master server to run the 'syncui' program. Note that it might be useful to capture the output of these commands. Refer to Setting Up an Interactive Command Line Interface for SYNCUI for tips specific to your platform.

Also, the DPX schedule is handled separately from the DPX enterprise configuration and catalog of jobs. It may be useful to review Using the SYNCUI Utility to Display an Existing Schedule as well for additional information on schedule related issue.

Once you have the appropriate command line interface set up, run the following command:

syncui

The syncui program will start and wait for interactive input. You will need to know the master server host name or IP address, and the DPX sysadmin password. Replace MSERVER below with the master server host name or IP address, and replace PASSWORD with the DPX sysadmin password. Enter the following syncui commands. These commands do not change anything in the database. After each command, take note of the "rc" return code. Specific instructions for analysis will follow below:

c s MSERVER

db login sysadmin PASSWORD

sep ,

db repair,catalog, , ,check,

device list

node list

group list

map list

tvol list

cat list

quit

And then take the DPX master down again for analysis. You will need to check the following:

 

db repair

If the'db repair' returns a non-zero return code, then there is some error that needs to be reviewed by Technical Support. If 'db repair' returns rc=0, then you will need to perform an additional check by reviewing the most recently written ssdb.*.log file. From this log look for the following message:

Processing message REPAIR CATALOG

From there, review the log messages and look for any RC codes that are non-zero, db files that fail to open properly, and any messages of the form:

Invalid file descriptor

If any of these error messages exist there may be a problem that needs review. Continue to review the ssdb log until you see the final message:

Message REPAIR CATALOG processing completed

 

list commands

Each of the list commands should return status code rc=0. Please look at the output and check that it looks sensible. 'device list' may not print any entries if you don't have any tape or diskdirectory devices. 'map list' may not produce any output if you do not run Archive to tape or Tape Migration jobs. If any of the indicated 'list' commands do not produce any output where you were expecting output, there may be an issue that needs to be followed up on. Please note, the 'tvol list' and 'cat list' commands can produce a lot of output that may take some time to print.

 

Follow-up

Any suspected errors should be reviewed by DPX technical support. It might be necessary to run similar checks on the old/original master server and DB files to see if the data base issues pre-existed the catalog move, or are suspected to be a result of the catalog move.