I have some servers that the agent stops checking in after the Audit Script starts? Maybe this question has been asked before? Does anyone know the cause of the agent going offline during "Latest Audit"?
YES, kaseya seems to think it's an issue with the iis persistant tcp timeout session. they recommend to set it to 2 hours !!!! I'm still testing this solution as I too so I'll let you know how it goes. :(
so far this is working. I've not gotten any 'false positives' on agent off-lines. I've had my tcp timeout set to 2 hours now for about a week and haven't had any servers drop.
Where and how do you set this value?
Do anyone have the configs for IIS 7.5?
IIS manager - Webistes - Explode default website - in Action Menu right pane - Advanced Settings - Behavior - Connection Limits - Connection Time-out (seconds) Change value 120 to 7200 - OK
Thank you......that's exactly what I did. Anything I need to adjust on the ASP side??
i did not change anything in ASP section, works fine ... no more false alerts..
I tried this, followed your steps exactly and I still have a server go offline randomly, it does have to be doing any audit. Also I am able to VPN to the site and can see the agent is running and the pushpin (that is the icon we use for the agent) is green indicating a good connection to the KServer. If I restart the Kaseya Agent service it brings it back online.
When this first started happening for us we would uninstall restart and reinstall the agent, I have since discovered that restarting the service works.
Has anyone else had this issue? If so is there anything else other than the step shown above?
I have a couple of servers were the agent stops responding. We jill the KAUSERTSK process and l click on the EXE to start it again.
From: gbath [mailto:firstname.lastname@example.org] Sent: Friday, June 17, 2011 12:32 PMTo: email@example.comSubject: Re: [Agent Discussions] Agent goes offline
it's not an issue with the agent, the agent is working fine. check for performance bottle necks. you can also run a report on the server 'event log frequency' for say the past 7 days, is that server overloading your I/O for the database? If the server receives more then 1000 events within an hour, it ignores those connections for the rest of the hour. I don't know if any of this will help, but it certainly can't hurt.
I've found (when trouble shooting this previously) that cleaning up the noise on the server helps a lot as well. If you or the customer is not willing to take the steps needed to clean up the event logs you can set them in the blacklist xml file to tell the agent to stop sending those specific alerts to Kserver.
@ John, I think that's a bad habit to fall into. You shouldn't have to restart the agent to keep it running, and if you look, the agent is not having any trouble, it's the server not reporting agent status correctly. The icon shows connected, the service doesn't stop running. I've had Kaseya engineers from support in my system looking at the agent and they've seen that the agent is acting as intended and reporting and keeping the connection. The issue is with the kserver.
I found out what my issue is, the server is getting spammed with WOL requests from another server, it happens every 2 seconds so I gues this would cause the agent issues trying to check in.