Need a bit of help. Sometime my Action can take 30-45 minutes before they start Evaluating, Downloading and eventually Running in the Console? Some jobs I need to turn around pretty quickly, is there a setting in the console that will allow my jobs to IMMEDIATELY begin processing? A lot of my work is time sensitive and I need a solution for this. Any help would be appreciated.
Thanks
Sno
When you execute a task (that is if if you are allowing UDP to 52311 to that endpoint) a udp packet will get sent to the endpoint to tell the client that it has a task to process. If you are not allowing UDP because it’s in a DMZ or public internet scenario then it will relay in the normal client heartbeat or Command polling to check in and process events. If your not allowing UDP on 52311 (or quite frankly even if you are) I would recommend looking at command polling if you are not already.
Outside of that, Actions can take longer to process because of long evaluation cycles or long processing relevance within your task.
It’s also worth noting that the task is always going to evaluate locally on the client, long before you see it show evaluating in the console, but 30-45 minutes sounds more like UDP isn’t allowed and it’s relying on other methods to check in.
Completely agree with @Jstev, make sure UDP is working.
If you have to use polling, be careful, every time your “poll” occurs, it interrupts and causes the evaluation to restart. You polling time should always be longer than the average evaluation time on a machine.
To determine evaluation times…
Maximum
((maximum of it /1000* second)/minute) of evaluationcycle of client
Average
((average of it /1000* second)/minute) of evaluationcycle of client
Top 10 longest Evaluation Fixlets
track fixlets of evaluationcycle of client
To set polling to 10 minute
setting "_BESClient_Comm_CommandPollEnable"="1" on "{parameter "action issue date" of action}" for client
setting "_BESClient_Comm_CommandPollIntervalSeconds"="600" on "{parameter "action issue date" of action}" for client
All good points here. One other potential consideration here that may help is Persistent Connections in cases where notifications via UDP are challenging.