I'm running MWG 7.0, build 7.0.2 (9231) and users keep getting random Bad Gateway errors. Has anyone else experienced this or provide any direction on where to start in trying to resolve this?
This thread is unrelated but you can follow the steps for troubleshooting.
If you can replicate and get it on a network capture we can look at it for you.
I resolved the issue by increasing the Connection Server Timeout. Configuration > Proxies (HTTP, FTP, HTTPS). I started by increasing the value to 60 but a few sites still had issues. I have increased it to 90 seconds and have not had the issue since. The only downside to the increase in this value is that, if a website fails, you won't know until after 90 seconds.
I noticed this to be a problem especially for, example, travel sites that leave a users "waiting" at a processing type screen while it pulls the results of a travel querey, etc.
Hope this helps,
just as a note. the setting you mentioned has increased the global timeout. This can be ok, but can also have side efects! For websites which sho this issue, you can put their names into a list and apply a workaound to them. The workaround setting allow you to apply this to a subset then only rather than chaning the global value.
just as a thought.
Thanks Michael. Is there any documentation on how to create the subset timeout value and apply a list to it?
I have had the setting changed now for about a month and have not noticed any negative side effects from increasing the value. (3,000+ user load) I am wondering though, the issue only came up when we migrated to the 7.x software. What is the value for this setting in the 6.8 software?
Awesome. I have been troubleshooting an issue with one incredibly slow site and the suggestion of increasing the connection timeout came up, but I didn't want to make that change globally.
I've imported your ruleset and will test with the customers ASAP.
I added the ruleset (and applied it to both requests and responses) and I still receive "Bad Gateway" errors on that site.
I do have a ticket open on this issue and will report to the engineer the steps taken. (They were apparently unaware of the ability to selectively modify timeouts - but then again, it's not working yet, so.... )
Thanks for the tip. I really do appreciate it.
I have not had success with the individual ruleset. If I resort the global setting back to its default and enter the websites exhibiting the problem into the ruleset, the problem persists. If I adjust the global setting, the problem resolves. Seems to vary by experience though. Thought I'd share.