Problem with nexus server migration to new hardware


#1

Hello
I’m using OSS 3.12.0-01 and trying to migrate from old server (aix 6.1) to new server (CentOS Linux).

I’ve copied all my data folders (whole sonatype-work) on new hardware while old server was stopped, but
every time I’m trying to start new server there are problems with orientdb:

*2019-04-10 08:20:49,907+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Magic number verification failed for page ‘0’ of ‘internal.pcl’.

I’ve tried remove all old orientdb and restore them from backup as written in documentation but errors with magic number still exists:

*2019-04-10 08:20:46,125+0200 INFO [FelixStartLevel] SYSTEM ROOT - Storage ‘plocal:/nexus/sonatype-work/nexus3/db/config’ is created under OrientDB distribution : 2.2.34 (build f340442755a31eabc91b87cb3ef99eda5cee6ebd, branch 2.2.x)
*2019-04-10 08:20:46,134+0200 INFO [ForkJoinPool.commonPool-worker-2] SYSTEM ROOT - Storage ‘plocal:/nexus/sonatype-work/nexus3/db/security’ is created under OrientDB distribution : 2.2.34 (build f340442755a31eabc91b87cb3ef99eda5cee6ebd, branch 2.2.x)
*2019-04-10 08:20:49,606+0200 INFO [FelixStartLevel] SYSTEM org.sonatype.nexus.orient.internal.DatabaseRestorerImpl - restoration of database config from file /nexus/sonatype-work/nexus3/restore-from-backup/config-2019-04-07-10-31-52-3.12.0-01.bak starting
*2019-04-10 08:20:49,606+0200 INFO [ForkJoinPool.commonPool-worker-1] SYSTEM org.sonatype.nexus.orient.internal.DatabaseRestorerImpl - restoration of database analytics from file /nexus/sonatype-work/nexus3/restore-from-backup/analytics-2019-04-07-10-31-52-3.12.0-01.bak starting
*2019-04-10 08:20:49,606+0200 INFO [ForkJoinPool.commonPool-worker-3] SYSTEM org.sonatype.nexus.orient.internal.DatabaseRestorerImpl - restoration of database accesslog from file /nexus/sonatype-work/nexus3/restore-from-backup/accesslog-2019-04-07-10-31-52-3.12.0-01.bak starting
*2019-04-10 08:20:49,854+0200 WARN [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Page 0 is broken in WAL segment 0 and will be truncated
*2019-04-10 08:20:49,855+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} 1 pages in WAL segment analytics.31.wal are broken and will be truncated, some data will be lost after restore.
*2019-04-10 08:20:49,875+0200 WARN [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Storage ‘analytics’ was not closed properly. Will try to recover from write ahead log…
*2019-04-10 08:20:49,876+0200 WARN [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Restore is not possible because write ahead log is empty.
*2019-04-10 08:20:49,901+0200 INFO [ForkJoinPool.commonPool-worker-2] SYSTEM org.sonatype.nexus.orient.internal.DatabaseRestorerImpl - restoration of database security from file /nexus/sonatype-work/nexus3/restore-from-backup/security-2019-04-07-10-31-52-3.12.0-01.bak starting
*2019-04-10 08:20:49,906+0200 INFO [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Storage data recover was completed
*2019-04-10 08:20:49,907+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Magic number verification failed for page ‘0’ of ‘internal.pcl’.
*2019-04-10 08:20:49,908+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Magic number verification failed for page ‘0’ of ‘index.pcl’.
*2019-04-10 08:20:49,909+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Magic number verification failed for page ‘0’ of ‘manindex.pcl’.
*2019-04-10 08:20:49,909+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Magic number verification failed for page ‘0’ of ‘default.pcl’.
*2019-04-10 08:20:49,909+0200 ERROR [ForkJoinPool.commonPool-worker-1] SYSTEM ROOT - $ANSI{green {db=analytics}} Magic number verification failed for page ‘0’ of ‘orole.pcl’.

I though that something is wrong with my Linux machine, so I’ve tried set up nexus server on windows pc. Unfortunately the same error with magic number as soon as restore complete.

Any advices ?


#2

The database looks to be corrupt. Does it boot on the original system? If so, I suggest trying the copy again, something went wrong during this attempt.

Rich


#3

It is boot on original system. There were some problems on the original system (unexpected shutdown, database in ready only mode, but I’ve fixed this with rebuilding indexes), but right now original system works, but there are some errors from time to time about rolling back transactions.

I’ve tried copy this several times, using different tools (rsync, scp), the same result. Moreover the same magic number error appears from copying all nexus directories, or restoring database from backup (original system complete backup with success). I’ve even tried to export/import database using orientdb console. The same.

Am I missing something obvious here ? I’s moving nexus between different systems (unix, linux) allowed ?


#4

I’ve tried remove all old orientdb and restore them from backup as written in documentation but errors with magic number still exists:

There is no need to do that. Just copy the work directory while Nexus is not running. See here for more information: