This message was deleted.

Weird, ip or fqdn?

Indeed... DNS seems ok, its currently on FQDN, which matches prod. I'll try on the IP and see what happens, cheers Leee

Failing that, perhaps setup a load balancer and see what happens to that in the logs. Don't put it in as an STA, just to see if it exhibits the same behaviour. If it does, I'd be looking at event logs on the controllers/cloud connectors.

I would add LB services pointing at the STAs and keep an eye on them. I would check the services on the DDC at time of issue and you could run a packet capture to make sure there isn't something funny occurring like dodgy networking where it is sending traffic from the wrong snip ect

Nice one - cheers lads 🙂

I have seen weird firewall timeout issues sometimes - you can try adjusting the default 2 min retry interval to something more frequent as per https://support.citrix.com/article/CTX231916

Lads, this one was put on hold for a while, but the client has re-engaged. They are trying to get me linked up with one of their network engineers. I've tried changing the polling intervals for the STA service, no joy. My fake/test LB is coming up and going down at the same time as the STAs... the LB just reports TCP timeout. Once I reboot the netscaler the LB and STAs come back to life. After more testing, I see they are all going down at the same time within 2 hours. Ive also added an existing prod DDC/STA, which does the same... The Netscaler is hosted on VMware, the infra tech says that the networking, VLANs/Nics etc are identical to prod. Both prod and the new do not have MBF enabled (or client side keep alive), any other ideas? Hoping to get on a call with the network engineer before the break, cheers

I would look to move the new VPX onto the same ESXI host as prod and see if you get the same symptoms. It sounds VMware networking related to me I would look at the vswitch on the host and look at the vmnics used as up links are they set to correct speed and duplex ? Is the MTU set correctly on the vswitch ? Teaming set correctly? Any traffic shaping ? You could also get the network team to check the switch ports that the physical NICs are connected to for any CRC errors and discards ect

Nice one @c4rm0 - Great ideas 🙂

Hey John, an idea. Stick a new Ubuntu vm or windows (whatever) on the same host and run Test-NetConnection continuously and log it out. See if you get a similar sort of timeout int he same period. You'll only want a TCP connect same as the NS and to skip the ping. Let me know if that doesn't make sense.

Thanks @Leee Jeffries - Makes sense. I'll keep you posted 🙂

Oh..... yeah..... that'll do it. Lol