Heartbeat Setting

Usage and Config Platform
1

(imported topic written by rmnetops91)

Question,

Can anyone explain what the heartbeat setting does?

We have some machines that go offline in our console (gray), but are still pingable. Our polling in the masthead is set to every 24 hours, so if a machine doesn’t have any new actions, they may not check for new actions on their own for more than 24 hours. Would this cause the machines to show offline after 45 minutes (which is what our console “offline after” setting is set to)? If we send the machine’s a refresh, they will update in the console with the current last report time.

Shouldn’t machines only show offline in the console if they are truely “offline”? That doesn’t seem to be the case. We verified they talk with the relays fine, and they receive refresh UDP packet commands from the server.

2

(imported comment written by SystemAdmin)

See this thread to see if it may be the same or similar issue.

http://forum.bigfix.com/viewtopic.php?id=4065

3

(imported comment written by rmnetops91)

Nope. It’s not related to a specific OS. I was more curious how the Heartbeat worked (what it’s function is).

4

(imported comment written by BenKus)

Hey rmetops,

The “heartbeat” setting is the default time at which the computer checks in by reporting its latest property differences… You also have the “Mark offline after…” setting that will control when the computers get “grayed out” in the console… If your “Mark offline after” is too low, then you will see computers grayed out when they shouldn’t be…

What is your heartbeat currently set to?

Ben

5

(imported comment written by rmnetops91)

Send heartbeat is every 15 minutes. Mark offline is at 45 minutes.

Machines are set in the masthead to report in at least every 1 day if they receive no actions or refreshes.

6

(imported comment written by BenKus)

I think your agents are busy for some reason and so they skip reports to avoid using too much CPU… Try changing your heartbeat to 20 minutes (or longer) and your “mark online after” to 60 min (or longer) and see if that helps…

You might want to talk to support or get an optimization service from our professional services team:

http://support.bigfix.com/services/

Ben

7

The heartbeat setting is set in the console Preference dialog. The console uses a simple rule to mark computers as off-line; if a computer has not reported within 3 times the heartbeat interval, it is grayed out.

The platform communicates the heartbeat interval to the agents in an analysis in the action site. For example, when the heartbeat is set to 1 hour, the analysis looks like this:

  Subject: Heartbeat Analysis
  X-Relevance-Evaluation-Period: 01:00:00
  X-Fixlet-ID: 2
  Content-Type: application/x-bigfix-itclient-analysis

q2: apparent registration server time

The agent content evaluation logic uses the X-Relevance-Evaluation-Period value to delay evaluation. When this logic encounters an analysis whose evaluation period has lapsed, it marks the analysis for re-sampling during the next report which, if nothing else triggers a report, will occur when the agent finishes a round of evaluating all the content.

When a machine is showing off-line it can mean there is too much content for the agent to evaluate within the heartbeat interval. The agent limits its cpu usage, so increasing the configured usage could fix the problem. You can also use the ‘track fixlets’ inspector to identify content that is extending the duration of the content evaluation loop.