[DUAS] Dynamic razfic after an outage on a File System
search cancel

[DUAS] Dynamic razfic after an outage on a File System


Article ID: 191735


Updated On:


CA Automic Dollar Universe


After facing an outage (file system issue), no Uprocs could run at all.
The Launcher & Calculator are not working as expected.


Release : 6.x



Because the file system faced an outage, many files were corrupted.



Please do not attempt to perform the below procedure unless you have first attempted to fix the issue by recreating the objects / performing an offline reorganization / disabling the Tasks and some other first troubleshooting steps.

The consequence of this procedure will be the following:

Loss of previous launches in the Job Runs menu
Loss of Job Run History
Loss of awaited Events from / to other nodes
Loss of the Audit Trail and Reporter Statistics
Loss of the History Trace of all previous executions
Tasks will calculate the next launch according to the current time.

Perform a dynamic razfic as described below (replace X by the necessary area like A,I,S):

First of all take a copy of all jobs in Launch Wait and Running and Event Wait statuses into an Excel file to remember what jobs were scheduled or had to run.

- Login as the Dollar Universe administrator user on the target server and load the Dollar Universe environment

. ./unienv.ksh

- stop the node

- Perform a dynamic razfic of the exchanger, launcher, history trace, audit trail, awaited resources and statistics:

$UNI_DIR_EXEC/uxrazfic u_fecd60 X
$UNI_DIR_EXEC/uxrazfic u_fecl60 X
$UNI_DIR_EXEC/uxrazfic u_fmev60 X
$UNI_DIR_EXEC/uxrazfic u_fmsb60 X
$UNI_DIR_EXEC/uxrazfic u_fmsp60 X
$UNI_DIR_EXEC/uxrazfic u_fmtp60 X
$UNI_DIR_EXEC/uxrazfic u_fmhs60 X
$UNI_DIR_EXEC/uxrazfic u_fmph60 X
$UNI_DIR_EXEC/uxrazfic u_fmat60 X
$UNI_DIR_EXEC/uxrazfic u_frrv60 X
$UNI_DIR_EXEC/uxrazfic u_fmhr60 X
$UNI_DIR_EXEC/uxrazfic u_fmcx60 X
$UNI_DIR_EXEC/uxrazfic u_fmfu60 X
$UNI_DIR_EXEC/uxrazfic u_fmer60 X
$UNI_DIR_EXEC/uxrazfic u_fmeq60

- launch a reorganization offline:

- start the node in maintenance mode (starts only the IO)
$UNI_DIR_EXEC/unistart -m 

- Generate the launches for Task without sessions
$UNI_DIR_EXEC/uxupd tsk tsk=* 

- Generate the launch for Task with sessions
$UNI_DIR_EXEC/uxupd tsk tsk=* ses=*

- Through the UVC console  start the CDJ and the Calculator (Jobs should be visible in Launch_wait for now)
- Through the UVC console - Server Engine List start all the required Engines (EEP, DQM, Exchanger, Supervisor...)
- Now compare the Launches that are missing from the Excel taken at the time the node was stopped and add via UVC or script any manual launches needed and jobs in event wait that have now disappeared
- Through the UVC console - Server Engine List start now the Launcher


Additional Information


After the razfic of u_fmsb60 file, the tasks will calculate the next launch according to the current time.

To be sure that that all of the Tasks have a valid date you can execute the two below commands that any of the tasks
not miss its launch:

- uxupd tsk tsk=* (generate the launch for Task without sessions) 

- uxupd tsk tsk=* ses=* (generate the launch for Task with sessions)

If this is ocurring on a Windows OS, please replace the / by \ in the commands and $UNI_DIR_EXEC by %UNI_DIR_EXEC%