search cancel

Post-upgrade to UIM 20.4 AC loads slowly, or times out, nas crashes and data_engine queues

book

Article ID: 241739

calendar_today

Updated On:

Products

DX Unified Infrastructure Management (Nimsoft / UIM) DX Unified Infrastructure Management (Nimsoft / UIM) CA Unified Infrastructure Management SaaS (Nimsoft / UIM) Unified Infrastructure Management for Mainframe

Issue/Introduction

Upgraded UIM/Nimsoft from version 9.0.2 to 20.4.

- After upgrade, Admin Console page loads very slowly and gives a request timeout most of the time.

We have also increased java initial time but the page still loads slowly. Load page time: 50 sec.

- NAS is crashing over and over.

- data_engine messages remain queued:

May 14 14:20:37:627 [14168] 1 de: Creating bulk object - table: RN_QOS_DATA_0048 for qos: QOS_ORACLE_SORT_RATIO 
May 14 14:20:38:084 [4580] 3 de: Database_global_lock LOCK 
May 14 14:20:38:084 [4580] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:084 [4580] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:087 [15148] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [12724] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [13632] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [15148] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:087 [15948] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [4784] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [4068] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [7580] 3 de: Database_global_lock LOCK 
May 14 14:20:38:087 [15148] 1 de: monitor_thread - exit worker thread 
May 14 14:20:38:087 [12724] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:089 [12724] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:089 [13632] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:089 [13632] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:089 [15948] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:089 [15948] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:089 [4784] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:089 [4784] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:089 [4068] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:089 [4068] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:089 [7580] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:089 [7580] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:111 [9856] 3 de: Database_global_lock LOCK 
May 14 14:20:38:111 [9856] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:111 [9856] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:112 [14172] 3 de: Database_global_lock LOCK 
May 14 14:20:38:112 [14172] 3 de: Database_global_lock UNLOCK 
May 14 14:20:38:112 [14172] 1 de: commit_thread - exit worker thread 
May 14 14:20:38:161 [11876] 0 de: qos_data_thread - stop received ... 
May 14 14:20:38:161 [11876] 3 de: Before WaitForSingleObject 
May 14 14:20:38:290 [3424] 3 de: [LSV] has disconnected from database

Cause

- data_engine restarting/connectivity intermittent, plus leftover wasp webapp artifacts.

Environment

  • Release : 20.4
  • Component : UIM NAS
  • wasp 20.40
  • Oracle database

Resolution

1. Admin Console slowness

    Resolved as a result of the resolution for the data_engine which was frequently restarting as well as deleting leftover webapp artifacts from the Primary hub wasp which are no longer used/valid.

2. NAS crashing

  • This is a known issue with LUA scripts in nas v9.34 which is resolved in nas v9.34HF1. Once this was applied the nas crashing subsided.

   Release notes:

  • Fixed the crash issue due to Lua Script  (Upgrade on probe version 9.34)
  • Removed the 'override' from cfx which can be seen in fresh installation only

3. data_engine queuing

  • data_engine queueing was resolved by manually running the data_engine sqlserver_update_qos_type.sql script in MS SQL Server Studio so we could see a successful result, since based on the data_engine log entries, the data_engine seemed to stall when running that specific script, but would finish without error after some time. The script would finish executing, but the delay seemed 'unusual' so manually running it again and ensuring it was successful fixed the issue. The script is locate din the data_engine folder in the scrips directory.