BigIP Report Old

v5.4.0 tries to talk to all the configured f5s in parallel. Running over 100 in one xml will put a serious load on the box. I'll look into putting a process limit on it. In theory we can start some scripts as the other scripts end and get better performance out of it. I've run against 34 f5s and the old script doing them in series took over a half hour to run. The new one runs in less than 15 minutes, but with much higher cpu load on the box running the script.

I run the build script over vpn against all our f5s that are remote without issues. We have a number of different datacenters including an instance at Amazon and I've had no issues with the WAN.

With v14 an v15 I have had issues with REST (and ssh) logins.  mentions above the tmsh commands to alter the ldap timeout. I'm testing that now on our development systems.

I'm still running both the release and beta versions, one on the half hour, and one on the hour. They both seem to run fine. My live F5s are v13 through v15 systems.

Interested to hear how 100 systems scales. Does the box you run the script on have many cores? It should take advantage of that. Worst case you may have to split them up and have multiple versions of the report until I put some process throttling in place.

Keep me posted!