[f-nsp] Problems with serveriron 10.2.0d code

[Alex Sharaz]

Hi,
We've been running ServerIronXL hardware for a few years and I've just purchased some new chassis based replacements with WSM modules.

I've been porting our existing load balanced services over to the new boxes and have come across a funny.

I provide a server load balanced DNS service via our foundry kit. and have 3 "remote" servers defined and bound to a VIP. I've configured the serveriron to perform a zone health check and on each of the 3 servers I can see the health check coming in.  The same real servers are also used by our XL boxes in the interim.

What I've found is that weven though the serveriron thinks all 3 servers are active, when it tries to use the last one in the group of remote servers the query never reaches the server. a sh server real doesn't show any connections to the server. From the users point of view the query never resolves. If i take down the dns server on the remote server so that the serveriron health check fails, it succesfully uses the other two servers and everything is o.k. 

I've switched off iptables on my real(remote) servers but it doesn't make any difference. As far as i can see all my servers are identical. I have the same setup on another serverion on our west campus with anoter 3 servers and it does the same thing.

Anyone else seen this?

TIA
Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/foundry-nsp/attachments/20080805/b1c116dd/attachment.html>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://puck.nether.net/pipermail/foundry-nsp/attachments/20080805/b1c116dd/attachment.ksh>