Duplicate Export Process after upgrade to 9.0 (230678)

Duplicate export readers seen after upgrading shareplex from 8.6.6 to 9.0.0

Errors seen in event log

Error    2017-06-21 03:35:47.486829 10460 1527142928 Export: No such file or directory rmu_RSopen /home/anu/866/var866/downgrd/RS.0a01174f.6 failed  (exporting to alvsupu15.prod.quest.corp queue alvsupl14) [module rim]
Error    2017-06-21 03:35:47.487164 10460 1527142928 Export: Cannot downgrade routing spec file for 80606064 target   (exporting to alvsupu15.prod.quest.corp queue alvsupl14) [module rim]
Error    2017-06-21 03:35:47.487415 10460 1527142928 Export: Unable to open routing specification file  (exporting to alvsupu15.prod.quest.corp queue alvsupl14) [module xpt]
Error    2017-06-21 03:35:47.487692 10460 1527142928 Export: Couldn't send routing information to import - exiting  (exporting to alvsupu15.prod.quest.corp queue alvsupl14) [module xpt]
Notice   2017-06-21 03:35:47.487962 10460 1527142928 Export: send routing spec to mport RS0a01174f.6 failed  (exporting to alvsupu15.prod.quest.corp queue alvsupl14) [module xpt]
Info     2017-06-21 03:35:47.489053 10457 2769904096 Export exited with code=1, pid = 10460  (exporting to alvsupu15.prod.quest.corp queue alvsupl14)
Info     2017-06-21 03:35:47.489851 18228 2769904096 Export launched, pid = 18228  (exporting to alvsupu15.prod.quest.corp queue alvsupl14)
Info     2017-06-21 03:35:49.924054 18228 27453968 Export connected to host on alvsupu15.prod.quest.corp
Info     2017-06-21 03:35:49.926129 18228 27453968 Export connected to import on alvsupu15.prod.quest.corp

6501904 Export: Error getting hfidx: Error getting address for 0x0a01174f-a01174f-a01174a: Name or service not known  (exporting to alvsupl14.prod.quest.corp queue alvsupl14) [module xpt]

Upgrade was done with new prod dir and temporary variable directory.  A new subdirectory called 'downgrd' is not created for existing old var dir as upgrade process will only create downgrd folder for the variable directory mentioned in prod dir/data/defaults.yaml file.

SOLUTION:

If you have multiple variable directories or installing prod dir in a different directory with a temporary variable directory, you must create 'downgrd' folder under the existing variable directory that you want to upgrade.  Then run ora_setup for each additional variable directory to upgrade.  For details, please refer to steps in SharePlex Upgrade Guide.

WORKAROUND:

1. Shutdown shareplex

2. go to  $SP_SYS_VARDIR and create subfolder 'downgrd'

mkdir downgrd

chmod 775

keep the same ownership and permission like the other subdirectories.

3. go to prod dir/bin and run qview to delete the extra export process

./qview -i

qview> qsetup <port#>

qview> list

The following queues exist:

 <srchost>+P+o.ORA11GR2-o.ORA11GR2

    subqueues range from 0 to 5

        WRITER  +PI+<sourcehost>+sp_mport+0x0a01179a  (target ip)

        READER  +PP+<sourcehost>+sp_opst_mt+o.ORA11GR2-o.ORA11GR2

  o.ORA11GR2+C

        WRITER  +PA+o.ORA11GR2+sp_ocap+o.ORA11GR2

        READER  +PR+o.ORA11GR2+sp_ordr+o.ORA11GR2

  <targethost>+X

        WRITER  +PR+o.ORA11GR2+sp_ordr+o.ORA11GR2

        READER  +PX+<targethost>+sp_xport+0x0a01174f-0000-0001  (exporting to <sourcehost>)

        READER  +PX+<targethost>+sp_xport+0x0a01174f-a01174f-a01179a  (source ip address) ----WRONG ONE to be deleted, it points to its own ip address

command is deluser <export queue name> <wrong export reader>

qview> deluser <targethost>+X +PX+<targethost>+sp_xport+0x0a01174f-a01174f-a01179a

qview> list

qview> exit

run ora_setup for this existing variable directory if you have not done so, then restart SharePlex sp_cop.

If you are not sure how to do this, contact support.