Home | Previous Page | Next Page   Migration to a Later Version of a Database Server > Preparing for Migration > Preparing for Migration >

Diagnostic Information That You Need Before Upgrading

Before you upgrade to a newer version of Dynamic Server, get the diagnostic information shown in Table 15. The information that you gather will be useful if you have problems or issues after the upgrade and need help from Technical Support. If you have problems, the information that you gather can be compared to information obtained after the upgrade.

You can print the pages containing Table 15 and then use the table as a checklist. After you get the information specified in each row, check the second column of the row.

Table 15. Checklist of Information to Get Before Upgrading
Information That You Need to Get Check Here to
Indicate You Have
the Information
Get the SQL query plans for all regularly used queries, especially complex queries, using SET EXPLAIN ON. ______Yes, I did this.
Run dbschema -d -hd for all critical tables.
The output contains distribution information.
______Yes, I did this.
Get oncheck -pr output that dumps all of the root reserved pages. ______Yes, I did this.
Make a copy of the ONCONFIG configuration file.

A copy of this file is useful because oncheck -pr does not dump all of the configuration parameters. In addition, a copy of the ONCONFIG file is essential if you need to revert to an earlier version of Dynamic Server.

______Yes, I did this.
Prepare a list of all the environment variables that are set using the env UNIX command. ______Yes, I did this.
During times of peak usage, obtain:

  • an online.log snippet, with some checkpoint durations in it
  • onstat -aF, -g all, and -g stk all
______Yes, I did this.
During times of peak usage, run the following onstat commands repeatedly with the -r repeat option for a period of about three to five minutes:

  • onstat -u, to see the total number of sqlexecs used
  • onstat -p, for read and write cache rates, to detect deadlocks and the number of sequential scans
  • onstat -g nta, a consolidated output of -g ntu, ntt, ntm and ntd
  • onstat -g nsc, -g nsd, and -g nss for the status of shared memory connections
  • onstat -P, -g tpf, and -g ppf
  • vmstat, iostat and sar, for cpu utilization

    Run these repeatedly for a period of about 3 to 5 minutes. Check pages for usage.

  • timex of all queries that you regularly run
______Yes, I did this.

Note:
For 32- to 64-bit engine upgrades, change SHMBASE and STACKSIZE according to the onconfig.std configuration file for the new version. Also, look in Table 17 or in the look in onconfig.std file for configuration parameters that are introduced with the new version of Dynamic Server.
Home | [ Top of Page | Previous Page | Next Page | Contents | Index ]