Backup and Restore

IDS11 Backup

Author: Doug Lawry

With the IDS11 backup to directory functionality backups have never been easier

Set the tape devices to a directory with BACKUP_FILTER set to gzip

ALARMPROGRAM.sh has BACKUP_COMMAND ontape -a -d. Daily archives via cron with ontape -s -L 0 -d. Zero scripting :-)

When does onbar read the onconfig file?

Author: Kernoal Stephens

Can anyone tell me whether a modification to the BAR_MAX_BACKUP parameter in the onconfig file requires the engine to be “bounced” for the change to take effect?

No the instance does not need to be bounced. Onbar read the onconfig file when it runs

Archecker

High Performance Loader

  • Create a project: onpladm create project foobar
  • Create a load job: create job ld_cust –p foobar -d <filename> –D tpch –t customer –fl –zD
  • View the job: onpladm describe job ld_cust –p foobar –fl
  • View the device: onpladm describe device ld_cust
  • Modify the device: onpladm modify object –F ld_cust_dev.job

Troubleshooting

Chunk fails to restore

If the restore fails with an unable to open chunk message and the chunk is a mirror then restore without mirrors.

Inconsistent Temp DBspace after restore

Some releases can experience problems restoring temporary dbspaces. This normally manifests itself as a downed chunk after the restore. This chunk can not be dropped because the engine still believes there is valid data contained within the chunk. Informix Technical Support are required to clear this problem.

If the engine can not be upgraded to avoid the problem, then drop the temporary dbspaces before the archive and put them back later.

Engine remains in Fast Recovery

If after the restore can complete the engine remains in fast recovery for a considerable period and the machine is idle stop and restart the engine, this can sometimes clear the problem.

If the engine supports it try setting the environment variable SINGLESEG_OFF to 1

Tapes not in the correct order

Informix will only restore if the tapes are presented in the correct order. However if the order is unknown then the tape sequence can read from the header of the tape, it should be at byte 15 and 16 on the tape.

dd if=<tape device> bs=1 skip=14 count=2 | od

It does fit on the tape

There are a number of reasons why the database won't fit on the tape. Probably the two most common are

  • The tape isn't really as long as it advertised.
  • There is compression in the way and you are not acheiving the compression rate that the manufacturer claims is possible. For example a 70GB DLT tape is in fact a 35GB tape with a 2:1 compression ratio. There are no hard and fast rules for estimating the amount of compression that will be seen.

Personal Tools