Jfreyre
Active Member
Can\'t build full packages on 8.95J1 due a Potext generation Issue
EnterpriseOne 8.11SP1 / JJ11556 / 8.95.J1 / Intel / Oracle 9.2.0.6.0
After we upgrade to 8.95J we are unable to build full packages from deployment server due a potext generation issue and R9621 never reach the end, after 17 hours trying to generate the spec files for potext on ClientPkgBuild.log we can see random BLOB records corrupt error, but when we open the generate potext file potext.ddb from the spec package folder with an hexadecimal editor and we search for the next registry on F98306 table we are unable to find such corrupted registries.
On the other hand when we try to build the same package from a workstation everything works fine but the generation of potext takes 15 hours, with a grow rate of 2 kb/s which is notoriously slower than gbrspec or fdaspec generation.
Re-generating index or rebuilding stats on F98306 didn't resolve the issue, but the data retrieval with UTB of thsi table on PY811, PD810 and Pristine is notoriously slower than gbrspec table. The F98306 table contains only 29 extent on the PY811's tablespace and the undo_retention parameter is set to 11 hours.
I suspect that the blob corruption message on ClientPkgBuild.log is produced due a long wait_time period so my goal is to reduce this unusual time consuming behavior.
Any clues?
Thanks for your answers.
J.A. Freyre.
------------------------------------------------
ClientPkgBuild.log
Fri Apr 28 22:17:56 - Building TAM and/or Pack file(s) for potext.
Fri Apr 28 22:17:57 - Spec count from Central Objects RDB: 63666
Sat Apr 29 15:09:28 - BLOB data for record number 48994 is corrupt. Correct the error. Spec Error: JDB operation failed!. Additional Error Code: 16
Sat Apr 29 15:09:28 - The record is: (Spec data is NULL.).
Sat Apr 29 15:09:28 - BLOB data for record number 48995 is corrupt. Correct the error. Spec Error: JDB operation failed!. Additional Error Code: 16
Sat Apr 29 15:09:28 - The record is: (Spec data is NULL.).
EnterpriseOne 8.11SP1 / JJ11556 / 8.95.J1 / Intel / Oracle 9.2.0.6.0
After we upgrade to 8.95J we are unable to build full packages from deployment server due a potext generation issue and R9621 never reach the end, after 17 hours trying to generate the spec files for potext on ClientPkgBuild.log we can see random BLOB records corrupt error, but when we open the generate potext file potext.ddb from the spec package folder with an hexadecimal editor and we search for the next registry on F98306 table we are unable to find such corrupted registries.
On the other hand when we try to build the same package from a workstation everything works fine but the generation of potext takes 15 hours, with a grow rate of 2 kb/s which is notoriously slower than gbrspec or fdaspec generation.
Re-generating index or rebuilding stats on F98306 didn't resolve the issue, but the data retrieval with UTB of thsi table on PY811, PD810 and Pristine is notoriously slower than gbrspec table. The F98306 table contains only 29 extent on the PY811's tablespace and the undo_retention parameter is set to 11 hours.
I suspect that the blob corruption message on ClientPkgBuild.log is produced due a long wait_time period so my goal is to reduce this unusual time consuming behavior.
Any clues?
Thanks for your answers.
J.A. Freyre.
------------------------------------------------
ClientPkgBuild.log
Fri Apr 28 22:17:56 - Building TAM and/or Pack file(s) for potext.
Fri Apr 28 22:17:57 - Spec count from Central Objects RDB: 63666
Sat Apr 29 15:09:28 - BLOB data for record number 48994 is corrupt. Correct the error. Spec Error: JDB operation failed!. Additional Error Code: 16
Sat Apr 29 15:09:28 - The record is: (Spec data is NULL.).
Sat Apr 29 15:09:28 - BLOB data for record number 48995 is corrupt. Correct the error. Spec Error: JDB operation failed!. Additional Error Code: 16
Sat Apr 29 15:09:28 - The record is: (Spec data is NULL.).