Temporary 503 and 404 for OpenShift hosted webapp

Question

I deployed a webapp to WildFly 8.0.0 on OpenShift. The application currently has very few users, but works fine. I use a free OpenShift account (I don't know if it is relevant) with a single cartidge, for WildFly.

Sometimes, when I access the application I get 503 (sometimes) or 404 (most of the times) errors.

It seems that I get these errors if the application has not been used for some time (something like 2 or 3 days). For about a minute, if I reload the page I get the same error. But after about a minute, I do not get the errors any more, instead the application is correctly available.

It looks like OpenShift "disables" webapps if they have not been used for some time, then "re-enables" them as required (but displays 503 or 404 during the "re-enabling" of the webapps).

=> Is this a bug? Is this a well-known issue of OpenShift?

=> How can I prevent my webapp from being unavailable?

Regards

Answer 1

On the free plan, OpenShift will idle any gears that have not received an external HTTP request or git push in 2 days, as per this FAQ.

If you are only using the 3 free gears, you can upgrade to the Bronze plan, which does not have any gear idling, and will not have any additional charges.

Answer 2

As diw stated, gear idling is part of the free plan and with the announcement of the bronze plan, you probably won't have to worry about that anymore.

However, if you want to stay on the free plan and if your application needs regular visits so you don't get those errors, you could setup a monitoring service (for example http://pingdom.com or http://uptimerobot.com) to check hourly and thus avoid having your gear idled. I found this out by accident when I moved a small personal site over to OpenShift and it never was idled due to the monitoring service hitting it.