Hi,
In an SNMP driver there is a polling manager table that triggers regular SNMP poll groups.
Due to a bad connection from a customer, some poll groups can fail while others succeed.
Now the customer wants to know per poll group if they were polled successfully or failed.
We could add a trigger on timeout on all parameters or convert it to SNMP GETs from QAction, but they’re not great solutions.
Would there be other ways to achieve this?
Hi,
The only way to know if a poll group failed is to add a trigger on a parameter with Time set to “timeout after retries”.
When a poll group contains single parameters it will depend on how to interpret this:
-If the group has multipleGet=”true”, then it will be sufficient to only provide a trigger on one parameter of that group because all data is requested through one SNMP Get, so if parameter fails then all parameters of that group will fail
-If the group has multipleGet=”false”, then it will be harder to interpret this because then each parameter is polled with an own SNMP Get message and triggers will be needed for every parameter in this group. The conclusion when 2 out of 4 parameters of that group are in timeout should then probably be flagged it as failed.
There will be a “flag” indicator needed. A trigger before group to set the flag to “Busy”. A trigger timeout after retries on the parameter(s) to set the flag as “Failed”. And a trigger after group to set the flag as “OK” in case the flag still had the “Busy” indication.
I see no possible alternatives. This will unfortunately require a lot of implementation work when there are a lot of parameters/groups to be polled and I wonder what the added value is going to be when this info is displayed what groups failed to be executed if it is already known that the connection is bad.
Regards,