Source:Troubleshooting Si3 Deduplication

From SEPsesam
Revision as of 15:37, 30 March 2022 by Unknown (talk | contribs) (Updating to match new version of source page)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Other languages:

Template:Copyright SEP AG en

Docs latest icon.png Welcome to the latest SEP sesam documentation version 4.4.3 Beefalo/5.0.0 Jaglion. For previous documentation version(s), check documentation archive.


Si3 Deduplication

Repairing Si3 deduplication store

Problem

  • There are problems with my Si3 deduplication store.

Solution

Repair the Si3 deduplication store as follows:

  1. Update the SEP sesam Server (where the Si3 runs) to the latest version, see Updating SEP sesam.
  2. Check the recovery log (in the Si3 root directory). If only one, two or three savesets are shown, check whether they have been deleted (purged) due to the expired EOL. If this is the case, delete the entire contents of recovery.log (but do not delete the file!). In a few seconds, the Si3 status should be changed to OK ; in the GUI and in the Web UI, the Si3 store will change from red to black. In this case the problem is solved. However, see also next step.
  3. If there are many entries in recovery.log, you have to use the repair command.
  4. Check if gc or fsck is running (sm_dedup_interface -d <Si3_drive> status). If they are, stop them with the command
  5. sm_dedup_interface -d <Si3_drive> gc stop and fsck stop
  6. Check again if 'gc or fsck are stopped (note that fsck can restart quickly).
  7. Start the quick repair; this only works if there are ddl files in <si3-root-dir>/trash/data.
  8. sm_dedup_interface -d <Si3_drive> repair start asc quick <si3-root-dir>/trash/data
  9. Use
  10. sm_dedup_interface -d <Si3_drive> repair start asc quick <si3-root-dir>/trash/data
    1. If the sanity state is not OK, use
    2. sm_dedup_interface -d <Si3_drive> repair start asc quick <si3-root-dir>/data/lostnfound
    3. If the sanity state is still not OK, use
    4. sm_dedup_interface -d <Si3drive> repair start asc quick <si3-root-dir>/data
  11. Check that the repair works:
  12. sm_dedup_interface -d <Si3_drive> status The following line confirms that it works: … GC: 0 FSCK: 0, Repair: 1
  13. Wait a few minutes. Check the repair with dedup-status. Once the repair is complete, the Si3 state should change to OK. If not, use FULL instead of QUICK repair. Have in mind that this can be quite time-consuming.

Si3 remains in "shutting down" state

Problem

  • Manually stopping Garbage Collection (GC) fails and consequently Si3 remains in the "shutting down" state.

Solution

  • Restart the Si3 daemon by using sm_main restart sds. For more details on stopping and starting the SEP sesam services, see How to Start and Stop SEP sesam.

Si3 deduplication may not work with NFSv4

Problem

  • Si3 deduplication may not work with Network File System version 4 (NFSv4).

Cause

  • SEP sesam operations, such as backup, restore and migration, may fail due to Java problems with NFSv4.

Solution

  • To avoid this problem, connect your backup devices via NFSv3.

Using sm_dedup_interface

4 4 3:Configuring and Administering Si3 Deduplication Store by using CLI/en

See also

Analyzing SEP sesam Log FilesHow to Start and Stop SEP sesamSi3 Deduplication