David, in that case please test 2 more things that doesnt work correctly either for me:
1. when the enhanced rsp is configured, it creates automatically an alarm policy as well. however the alarm policy for the rsp in the
plugin_metric.cfg contains "qos_target = ~.*", so in case that
A) either more rsp profiles are configured for that rsp or
B) the device/robot where the rsp probe is deployed has other probe as well, for instance cdm, this alarm policy rewrites the alarm (tresholds, etc) for other QoS as well.
2. when you create a group and configure an rsp profile that group and after that you add a device(s) into that group and because of some reason you have to remove a device from the group, the rsp continues monitoring the device even it is NOT in the rsp device group anymore.
so in my opinion, there are more problems with the enhanced rsp.
if you found the same problems in your lab, please let me know, whether you can get in touch with the dev/product team and ask them to fix them. thank you.
Original Message:
Sent: 05-07-2021 01:37 PM
From: David MICHEL
Subject: enhanced rsp 5.50 doesnt monitor all the chosen QoS
Yes found the problem in my lab, investigating.
------------------------------
Support Engineer
Broadcom
Original Message:
Sent: 05-03-2021 05:16 PM
From: David MICHEL
Subject: enhanced rsp 5.50 doesnt monitor all the chosen QoS
will set it up in my lab and test.
------------------------------
Support Engineer
Broadcom
NY
Original Message:
Sent: 05-03-2021 04:57 AM
From: Stefan Pivoda
Subject: enhanced rsp 5.50 doesnt monitor all the chosen QoS
does the rsp probe 5.50 configured via enhanced rsp mcs in OC 20.3.3 monitor and send all the configured metrics?
i deployed rsp 5.50 and configured it in OC via enhanced rsp mcs, but it doesnt display/monitor all the chosen Qos.
for example, from the cpu Qos it monitors and sends metrics only for QOS_CPU_USAGE and QOS_PROC_QUEUE_LEN although all these Qos was chosen to be montired in OC:
CPU Average Total Usage
CPU Average User Time
CPU Average System Time
Individual CPU Total Usage
Individual CPU User Time
Individual CPU System Time
Individual CPU Wait Time
in addition, QOS_PROC_QUEUE_LEN was NOT even chosen to be monitored, but it is monitored/sent/displayed in OC.
tried on 2 different UIM (20.3.3) instances and troubleshooted with dr nimbus, where i can see that only for QOS_CPU_USAGE and QOS_PROC_QUEUE_LEN is QOS_DEFINITION sent.
the plugin_metric.cfg contains "publish_qos = true" for all the metrics chosen in OC to be montiored and "publish_qos = false" for QOS_PROC_QUEUE_LEN (that is actaully monitored).
how does the rsp 5.50 work in your environment please? and if it works fine, do you have please any idea what could cause the issue in our? i havent touched the rsp probe gui ( = manual rsp configuration).
thank you,
stefan