HIPJ-1574

The registered job might not run

Phenomenon

For the HULFT IoT Manager in the Red Hat Enterprise Linux Server 7 environment, the registered job might not run. This problem causes the following wrong actions and influences your environment.

Wrong actions:

From Ver.1.4.0 to Ver.1.4.1

Commands that are used in the job that is specified by using [Job] of [To] on the Transfer Settings screen of HULFT IoT Manager might not be executed.

From Ver.1.1.0 to Ver.1.3.1

Commands that are used in the job that is registered in the [Job Information] of HULFT might not be executed.

From Ver.1.1.0 to Ver.1.4.1

If you are using the following functions of HULFT that is bundled with HULFT IoT, the commands that are used in the job might not be executed.

Pre-send Job

Post-send Job (Successful Job and Unsuccessful Job)

Remote Job Execution

File Trigger

Affected versions

From Ver.1.1.0 to Ver.1.4.1

Version when it was corrected

Ver.1.5.0

Affected module

HULFT IoT Manager

Condition to reproduce

This phenomenon occurs when all of the following conditions are met:

HULFT IoT Manager is introduced to Red Hat Enterprise Linux Server 7.
A job is registered by using HULFT IoT Manager or HULFT settings.
In the registration contents of a job or in a start job, an arbitrary command is registered by using the other than the absolute path specification.

Alternative workaround

Rewrite the HULFT service startup settings file like the following example:

Step 1. Create the hulftrcscript.sh file at the /opt/hulftiot-manager/hulft directory.

Example hulftrcscript.sh file
#!/bin/sh TARGET=$1 . /opt/hulftiot-manager/hulft/etc/hulft.bsh.profile hulcluster() { MODE=$1 hulcluster$TARGET -$MODE RET=$? if [ $RET -ne 0 ] ; then echo hulcluster$TARGET $MODE error. errcode=$RET return 1 fi return 0 } case "$2" in 'start') hulcluster start ;; 'stop') hulcluster stop ;; *) esac

Example hulftrcscript.sh file

#!/bin/sh
TARGET=$1
. /opt/hulftiot-manager/hulft/etc/hulft.bsh.profile
hulcluster() {
              MODE=$1
              hulcluster$TARGET -$MODE
              RET=$?
              if [ $RET -ne 0 ] ; then
                                  echo hulcluster$TARGET $MODE error. errcode=$RET
                                  return 1
              fi
              return 0
}
 
case "$2" in
'start')
              hulcluster start
              ;;
'stop')
              hulcluster stop
              ;;
*)
esac

Step 2. Correct the /etc/systemd/system/hulftrcv-iot.service file as follows.

Example hulftrcv-iot.service file
[Unit] Description=hulft8 daemon For hulclusterrcv After=network.target [Service] ExecStart=/opt/hulftiot-manager/hulft/hulftrcscript.sh rcv start ExecStop=/opt/hulftiot-manager/hulft/hulftrcscript.sh rcv stop Restart=no Type=forking [Install] WantedBy=multi-user.target

Example hulftrcv-iot.service file

[Unit]
Description=hulft8 daemon For hulclusterrcv
After=network.target
 
[Service]
ExecStart=/opt/hulftiot-manager/hulft/hulftrcscript.sh rcv start
ExecStop=/opt/hulftiot-manager/hulft/hulftrcscript.sh rcv stop
Restart=no
Type=forking
 
[Install]
WantedBy=multi-user.target

Step 3. If you are using the functions (Pre-send Job, Post-send Job (Successful Job and Unsuccessful Job), Remote Job Execution, and File Trigger) of HULFT that is bundled with HULFT IoT, you must correct the following files in the same way as above:

Correct the /etc/systemd/system/hulftsnd-iot.service file

Change the parts of "rcv" that are written in the correction example above to"snd".
Correct the /etc/systemd/system/hulftobs-iot.service file

Change the parts of "rcv" that are written in the correction example above to"obs".

Step 4. To reflect the contents of the service definition files (hulftxxx-iot.service) that are corrected by step 2. and step 3., execute "systemctl daemon-reload".

Status description

Even if the conditions described at the condition to reproduce are met, the registered jobs correctly run.

Official | HULFT IoT Release Note Fourth Edition: July 1, 2021

You are here: