Problem with INSSL Gateway

[rgilchrist]

Hi
We have an application with an INSSL Gateway connected to a Web server, which is running Centos 5 and Lighttpd 1.4.

We have noticed that web requests are being repeated on the web server, sometimes they are repeated up to 10 times.

It appears to be a problem with pound in the INSSL gateway. We are getting the following entries in the messages log from pound

"pound: error flush headers to xxx.xxx.xxx.xxx: Connection reset by peer"

And this results in numerous requests on the webserver.

It becomes a bit of a race condition with requests being repeated, causing the web server to slow down and this causes more repeating of requests and so on.

Just wondered if anybody had seen this before?


Cheers
Rob

[PavelGeorgiev]

Rob,

We stopped using pound in INSSL 2+ years ago because of multiple issues that it had and we strongly recommend upgrading to a newer version of INSSL/INSSLR (in newer versions of AppLogic INSSL is actually an assembly wrapper of INSSLR). If moving to a newer version of AppLogic is not an option, you can just upgrade that appliance.

INSSL(R) now uses nginx which is faster, more efficient and more reliable.

[rgilchrist]

Hi
Thanks for the reply, as a short term measure I have given the INSSL gateway more resources and that seems to have alleviated the problem, but I will definitely look into switching to the INSSLR gateways.

Cheers
Rob