Configuations caused error page

venu040 · 08-01-2009, 06:24 PM

Our Site went down one day by this message,



Please help me for investigation on this bad page

**wmorri** · 08-01-2009, 09:16 PM

Hi,

Two things I can think of right off the bat, one is to make sure that your index page is there and working. The other is can you explain why you are using Tomcat instead of just plain Apache?

Cheers!

venu040 · 08-01-2009, 10:45 PM

Hi thanks for looking into this!

Index page is working fine!
This issue was there only for a limited period of time and the bad page is not reproducible after flushing our third party Akamai cache, which is configured to cache the pages received from origin - directly hits our webservers).

We have been using tomcats for all application content which are not static, like .jsp, .do....etc actions. These tomcats are being called by apache webservers using ProxyPass.

User -> Akamai -> Loadbalancer -> Webservers -> Appservers.
If this makes sense/help.

I wonder when we do see this kind of error?

**wmorri** · 08-01-2009, 10:55 PM

Hi,

Thanks for the information. I would like to know what is happening with the server during the time period that your get the error. Is there a high load on the server, or high traffic. I am not familiar with Akamai, but do you have to flush the catche on it or can you just let the catche fill up for a while. I would have to think that if it only happens after you flush the catche then there has to be something in the catche or connected to the catche that is causing the problem.

I hope that this is making sense. I would have a look at the catche to see if there is anything that could cause an error in the servers. I might even have a look at the software help or talk to the company to see if they have heard of this isssue, as long at it doesn't cost you and arm and a leg.

Cheers!

venu040 · 08-01-2009, 11:10 PM

Hi,

Actually there wasn't any load or high traffic at the time of issue, we observed.
So, when we were able to open the home page of our site with origin, which means origin-www.site.com it hits our webserver directly without any problems. The only we could think of was Akamai cache occurs while the request was submitted. We contacted Akamai support as well and they end up saying that bad page was received and cached from origin, which is the domain hits our load balancer -> webservers as I explained previously.
In Akamai configuration we can cache any page with some TTL - TimeToLive. So, it was set up for 30mins. for homepage. The reason why we were sure of some bad page cached and made business discontinuity, quick resolution took place by clearing it manully without waiting for the actualy time limit completes.

But the error looks like some configuration changes I could think of... and its a commented out XML element... I wonder!

One more thing I wanted mention was OSCache, web.xml, was modified if incase that creates any issues. However issue was resolved before we reverted back the changes anyway!

08-01-2009, 06:24 PM	#1 (permalink)
venu040 Registered User Join Date: Aug 2009 Posts: 3 OS: Solaris 10 & Linux Red hat	Configuations caused error page Our Site went down one day by this message, <!-- fake default page for Apache so that it passes the request to Tomcat --> Please help me for investigation on this bad page

08-01-2009, 09:16 PM	#2 (permalink)
wmorri Moderator/Fedora Amb. Join Date: May 2008 Location: /pm/etc Posts: 2,835 OS: Window 7/Fedora 10 My System	Re: Configuations caused error page Hi, Two things I can think of right off the bat, one is to make sure that your index page is there and working. The other is can you explain why you are using Tomcat instead of just plain Apache? Cheers! __________________ Linux Forever!

08-01-2009, 10:45 PM	#3 (permalink)
venu040 Registered User Join Date: Aug 2009 Posts: 3 OS: Solaris 10 & Linux Red hat	Re: Configuations caused error page Hi thanks for looking into this! Index page is working fine! This issue was there only for a limited period of time and the bad page is not reproducible after flushing our third party Akamai cache, which is configured to cache the pages received from origin - directly hits our webservers). We have been using tomcats for all application content which are not static, like .jsp, .do....etc actions. These tomcats are being called by apache webservers using ProxyPass. User -> Akamai -> Loadbalancer -> Webservers -> Appservers. If this makes sense/help. I wonder when we do see this kind of error?

08-01-2009, 10:55 PM	#4 (permalink)
wmorri Moderator/Fedora Amb. Join Date: May 2008 Location: /pm/etc Posts: 2,835 OS: Window 7/Fedora 10 My System	Re: Configuations caused error page Hi, Thanks for the information. I would like to know what is happening with the server during the time period that your get the error. Is there a high load on the server, or high traffic. I am not familiar with Akamai, but do you have to flush the catche on it or can you just let the catche fill up for a while. I would have to think that if it only happens after you flush the catche then there has to be something in the catche or connected to the catche that is causing the problem. I hope that this is making sense. I would have a look at the catche to see if there is anything that could cause an error in the servers. I might even have a look at the software help or talk to the company to see if they have heard of this isssue, as long at it doesn't cost you and arm and a leg. Cheers! __________________ Linux Forever!

08-01-2009, 11:10 PM	#5 (permalink)
venu040 Registered User Join Date: Aug 2009 Posts: 3 OS: Solaris 10 & Linux Red hat	Re: Configuations caused error page Hi, Actually there wasn't any load or high traffic at the time of issue, we observed. So, when we were able to open the home page of our site with origin, which means origin-www.site.com it hits our webserver directly without any problems. The only we could think of was Akamai cache occurs while the request was submitted. We contacted Akamai support as well and they end up saying that bad page was received and cached from origin, which is the domain hits our load balancer -> webservers as I explained previously. In Akamai configuration we can cache any page with some TTL - TimeToLive. So, it was set up for 30mins. for homepage. The reason why we were sure of some bad page cached and made business discontinuity, quick resolution took place by clearing it manully without waiting for the actualy time limit completes. But the error looks like some configuration changes I could think of... and its a commented out XML element... I wonder! One more thing I wanted mention was OSCache, web.xml, was modified if incase that creates any issues. However issue was resolved before we reverted back the changes anyway!