I would like to only publish alarms for certain monitors in snmpcollector, without publishing data (qos).
This has always been possible in the past using the IM configuration of the probes.
Now with the Admin console probe configurator, if i deselect "publish data" i cannot set any static alarm for the monitor.
Is that working as designed? is there a workaround to this (other than setting the retention period to 1day for that monitor in data_engine)?
This is as working as designed.
SNMPcollector's alarm you are seeing is actually coming from baseline_engine which subscribes the value of QoS and publish alarm based on threshold.
snmpcollector Theory of Operations - CA Unified Infrastructure Management Probes - CA Technologies Documentation
QoS and alarm data is forwarded to the primary hub by two different processes. QoS data is passed from the remote hub to the primary hub through queues on the message bus. Queues must be configured for QoS data (QOS_MESSAGE and QOS_DEFINITION), and probe_discovery.
Every remote hub with snmpcollector must have baseline_engine, prediction_engine, alarm_enrichment, and NAS installed to enable threshold alarms. The baseline_engine probe monitors the message bus for QoS data that breaches the defined thresholds. The predictive_engine probe is required to enable TimeToThreshold settings, and the alarm_enrichment probe is required to enable TimeOverThreshold settings. As alarm messages occur, the data is passed on to the primary hub through NAS. Matching versions of NAS must exist on the primary hub and the remote hub so that alarm messages can be sent between the hubs. In order to avoid duplication of alarms, an alarm queue should not be present and the instance of NAS on the remote hub must be configured to send all alarm events in both directions.
The following diagram shows how QoS and alarm data flows from the remote hub to the primary hub.
I believe this is a very confusing behaviors to all UIM customers where they are used to IM configurable probes.
If you are seeing "Static Alarm" in your Admin Console GUI for a probe, this represents for threshold which is being managed by baseline_engine probe, not the monitoring probe itself.