When configuring a customized probe package, if you mistakenly put "interval" timespec for a probe definition while also specifying "daemon" type, the probe will restart repeatedly.
When you create a new probe you have to set the Probe Definition values for the controller.cfg, e.g. daemon probe or on_demand, etc. This is configured in Package Editor in Infrastructure Manager.
Normally when editing the probe options if you specify "daemon" probe type, then "timespec" field gets grayed out - you cannot change it or put anything in there which implies it should be ignored. (You can see this in IM by going to a probe which is running on some robot, right-click the probe and choose Edit....)
However, in some cases, when developing a custom probe package, it is possible to mistakenly set this to "Interval" (e.g. 3 minutes) and then once the probe is deployed, since it is grayed out, you cannot change it:
This will cause the probe in question to restart every 3 minutes, even though a "daemon" type probe should be ignoring this (since it is grayed out).
This will be corrected in the upcoming robot_update after 9.40 which will be included in the upcoming CU10/23.4 releases.
In the new version, the timespec field will be properly ignored when "daemon" type is selected.