Intermittent issues running umbraco in a load balanced environment (file lock + ResolveRequestCache)
Hi Guys
I went live with an Umbraco site yesterday and experienced several issues that are currently leaving me stumped.
First the set up.
2 Azure hosted VMs sitting in a load balancer using session persist. Media folder is being synced with a file sync utility both Umbraco front ends share the same database. Both VMs have sub domains pointing to each individual box and have SSL certs installed.
Both boxes have been tested for 3 days before going live with no issues.
The Issues
Once the site went live I came across the "Failed to enter the lock within timeout" error, this error was intermittent on both the VMs however they do not share the same filesystem. As a result the site would not load, however restarting the server fixed the issue initially but then intermittently came back. Ive been through this post: http://issues.umbraco.org/issue/U4-6626
However I have been unable to resolve.
The second issue I have discovered is an alarming about of failed requests When load testing the site I saw high levels of failures. In some cases 80%+ rate of failure suggesting the application isn’t scaling past low levels of traffic. This happens only on HTTP and not HTTPs. Upon checking new relic I can see something is happening with the ResolveRequestCache however I’m unsure what exactly.
Intermittent issues running umbraco in a load balanced environment (file lock + ResolveRequestCache)
Hi Guys
I went live with an Umbraco site yesterday and experienced several issues that are currently leaving me stumped.
First the set up.
2 Azure hosted VMs sitting in a load balancer using session persist. Media folder is being synced with a file sync utility both Umbraco front ends share the same database. Both VMs have sub domains pointing to each individual box and have SSL certs installed.
Both boxes have been tested for 3 days before going live with no issues.
The Issues
Once the site went live I came across the "Failed to enter the lock within timeout" error, this error was intermittent on both the VMs however they do not share the same filesystem. As a result the site would not load, however restarting the server fixed the issue initially but then intermittently came back. Ive been through this post: http://issues.umbraco.org/issue/U4-6626
However I have been unable to resolve.
The second issue I have discovered is an alarming about of failed requests When load testing the site I saw high levels of failures. In some cases 80%+ rate of failure suggesting the application isn’t scaling past low levels of traffic. This happens only on HTTP and not HTTPs. Upon checking new relic I can see something is happening with the ResolveRequestCache however I’m unsure what exactly.
I also saw this post : https://our.umbraco.org/forum/extending-umbraco-and-using-the-api/80538-cache-in-load-balanceed-environment
However I have been unable to resolve.
Many thanks
is working on a reply...