search cancel

Impossible to connect to AWI after ZDU following a dbload of the wrong version

book

Article ID: 253036

calendar_today

Updated On:

Products

CA Automic Workload Automation - Automation Engine

Issue/Introduction

On a TEST environment, an upgrade with zero downtime (ZDU) was performed from version 12.3.9 to 12.3.9-HF1, but it's impossible to connect via AWI afterwards, as AWI will throw a Timeout connection error.

History of actions performed:

    1. dbload UC_UPD.TXT  ( by mistake it was done with the wrong version 12.3.9)
    2. dbload UC_UPD.TXT  (with the correct version 12.3.9 HF1)
    3. New processes AE not recognized as up-to-date on step 2 ZDU
    4. Full stop of processes on both versions (this caused the total outage)
    5. Restart of the processes and since then AWI refuses to connect with the timeout Error

Investigation

On the AE logs, we can see the connection from AWI:

==> CPsrv_G01R02C03_003_00.log <==
20221018/114940.251 - U00003406 Client connection '12(6)'  from '[10.118.120.168]:53324' has logged on to the Server.
20221018/114940.318 - U00003459 Dialog '0,MIEC7426,ADMIN' logged on (Client connection='12').

==> WPsrv_G01R02C03_003_00.log <==
20221018/114940.579 - U00003205 Logon of ' ' ('MIEC7426/ADMIN'), client: '0000' accepted by host 'X.X.X.X'. Client version = '12.3.9+build.1655235965775', type='Java-API' (connection='*CP003#00000012').

But after three minutes ( default timeout from AWI), it times out and connections are closed:

==> CPsrv_G01R02C03_003_00.log <==
20221018/115244.547 - U00003407 Client connection '12(5)' from ''X.X.X.X':53324' has logged off from the Server.

==> WPsrv_G01R02C03_003_00.log <==
20221018/115244.572 - U00011852 Logoff ' ' ('MIEC7426/ADMIN'), client: '0000' (Connection='*CP003#00000012').

Environment

Release : ZDU from 12.3.9 to 12.3.9HF1

Component: Automation Engine

Cause

Duplicated (with same zdu_version) wrong entry on the table UC_ZDU with empty values for zdu_ucsjchngelist , zdu_cpchngelist and zdu_wpchngelist. This was caused by the first wrong dbload of the wrong version.

This could be found with this query:

select * from UC_ZDU;
 zdu_pk | zdu_mqset |           zdu_version           | zdu_active | zdu_base | zdu_ucsjchngelist | zdu_cpchngelist | zdu_jwpchngelist | zdu_wpchngelist
--------+-----------+---------------------------------+------------+----------+-------------------+-----------------+------------------+-----------------
      5 | 1         | 12.3.9+hf.1.build.1660287977073 |          1 |        1 | 1659686561        | 1656333657      | 1655199907       | 1656333657
      6 | 2         | 12.3.9+hf.1.build.1660287977073 |          0 |        0 |                   |                 | 1659170931       |

Resolution

To fix the issue, it was necessary to delete the extra entry on UC_ZDU with zdu_pk the one not active).

WARNING: DO NOT ATTEMPT THIS PROCEDURE WITHOUT THE APPROVAL OF TECHNICAL SUPPORT

Example of the procedure used:

  1. Backup the content of the table UC_ZDU or note the output of the following command to reinsert the deleted record in case of problem
  2. select * from UC_ZDU;

     zdu_pk | zdu_mqset |           zdu_version           | zdu_active | zdu_base | zdu_ucsjchngelist | zdu_cpchngelist | zdu_jwpchngelist | zdu_wpchngelist
    --------+-----------+---------------------------------+------------+----------+-------------------+-----------------+------------------+-----------------
          5 | 1         | 12.3.9+hf.1.build.1660287977073 |          1 |        1 | 1659686561        | 1656333657      | 1655199907       | 1656333657
          6 | 2         | 12.3.9+hf.1.build.1660287977073 |          0 |        0 |                   |                 | 1659170931       |

  3. Run the delete statement to delete the line with with some empty values for zdu_ucsjchngelist , zdu_cpchngelist, zdu_jwpchngelist and zdu_wpchngelist, this corresponds to zdu_pk=6 on the example above:
    delete from uc_zdu where zdu_pk=6;
  4. Run again the select statement to confirm that it correctly deleted the wrong record:
    select * from UC_ZDU;

     zdu_pk | zdu_mqset |           zdu_version           | zdu_active | zdu_base | zdu_ucsjchngelist | zdu_cpchngelist | zdu_jwpchngelist | zdu_wpchngelist
    --------+-----------+---------------------------------+------------+----------+-------------------+-----------------+------------------+-----------------
          5 | 1         | 12.3.9+hf.1.build.1660287977073 |          1 |        1 | 1659686561        | 1656333657      | 1659170931       | 1656333657

  5. Connect again via AWI, after having started the processes on the new version 12.3.9HF1, this time the connection will work fine.