BigFix Client reporting very slow

Anton 2021-01-21 09:44:36 UTC #1

Hello!

I have a problem with bigfix that client report very slowly.

I have 8 BigFix Servers under my control.
All of them a configured the same (from network, hardware and software perspective)

As should be (6/8 servers):
When a create new analysis (without any relevance, so all connected devices should be applicable) with 1 or more properties (lets say simple “operating system”) for 6 servers all computers (lets say 99% of them) become applicable for this new analysis in 5-10 min and show me its value.

Problem (2/6 servers):
When i create new analysis (without any relevance, so all connected devices should be applicable) with 1 or more properties (lets say simple “operating system”) for 2 servers it takes more that 48 hours for most of the computers to become applicable for this analysis and report its value to the server.

This also refers to creating just properties, assigning tasks, fixlets, automatic groups…
These 2 servers have very small amount of devices connected (100-200).

Please, help me to identify and solve the issue.

Best regards,
Anton Paradovskyi.

SLB 2021-01-21 10:22:24 UTC #2

I would check that clients can receive UDP 52311 from the relays to receive the “I have something for you” broadcast that is sent when you create or change fixlets, baselines, analyses etc.

If UDP 52311 isn’t open, clients fail over to polling for new content during a relay select, which defaults to every 6hrs unless you have configured your environment differently. If ICMP is also blocked, this will impede relay selection meaning clients then failover to a once every 24hr poll for content.

FatScottishGuy 2021-01-21 12:59:14 UTC #3

Exactly what Rob said though one point to note is to enable command polling.

You can configure that also by looking on here

https://help.hcltechsw.com/bigfix/9.5/platform/Platform/Installation/c_sample_proxy_configurations.html

Anton 2021-01-26 16:58:12 UTC #4

Hi SLB, FatScottishGuy.

Thank you for a help.

Issue is solved.
For 1 of the servers we were able to open UDP 52311.
For another network team refused to open it, so we applied manual relay selection on every 5 mins.

JasonWalker 2021-01-26 19:45:35 UTC #5

I’d suggest using either Persistent Connections, or Command Polling, or Both; but not to force a manual relay select every few minutes, as that puts additional workload upstream for client registration.

Let us know if you need a reference on these options or if you have any trouble finding the links.