Forum Discussion
NimbostratusBIG-IP LTM VE Trial Network Performance Problem
Hi All
I installed the BIG-IP LTM VE Trial (11.3.0 hotfix 1) using different versions of VMware (ESXi5.5, ESXi6 and Workstation 12) and am having the same network performance problem in all three cases.
I have unacceptable packet loss and latency between the LTM to other VM web servers on the same ESXi host / VM Workstation. E.g.
[root@BIGIP-1:Active] config ping 10.2.0.11
PING 10.2.0.11 (10.2.0.11) 56(84) bytes of data.
64 bytes from 10.2.0.11: icmp_seq=17 ttl=64 time=401 ms
64 bytes from 10.2.0.11: icmp_seq=18 ttl=64 time=526 ms
64 bytes from 10.2.0.11: icmp_seq=35 ttl=64 time=517 ms
64 bytes from 10.2.0.11: icmp_seq=47 ttl=64 time=441 ms
64 bytes from 10.2.0.11: icmp_seq=49 ttl=64 time=422 ms
Pinging between the host, web servers, and other VMs on the same network are fine and so I don't think there are any problems with ESXi.
Sometimes the response times are better (e.g. 20ms) but it is still high comparing to pinging between other VMs on the same ESX host.
I have been troubleshooting this for a couple of weeks now and it would be much appreciated if someone can shed some light on it.
Many thanks Domonic
2 Replies
Andrew_HuskingCirrus
The LTM VE trials and Lab editions i believe have a 10mb throughput limit.
if you tail the ltm logs you should see something along the lines of "10mb throughput limit hit", if that happens i believe it starts de-prioritising and dropping traffic.
Regards,
Domonic_Wong_25All fixed now.
The problem was caused by another VM which was running some security camera recording software and causing a lot of I/O which affected the network performance of the LTM VE somehow.
The ping response time was back to about 10-12ms after shutting down that VM.
Cheers Domonic
