I have a cloud service running on Azure and after hours/days it starts to get listed as unresponsive in the portal and visual studio server tab.
The websites that are hosted on the site still replies, but when its unresponsive its not included in the load balancing and therefore putting extra stress on other servers.
i started investigating but i am not sure where to go now. This is my findings from the c:/logs/
[00000006] [08/02/2014 16:38:38.57] [INFO] Role cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 is reporting state Ready. [00000010] [08/02/2014 16:38:39.58] [INFO] Role cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 has current state Started, desired state Started, and goal state execution status StartSucceeded. [00000006] [08/02/2014 16:38:43.62] [HEART] WindowsAzureGuestAgent Heartbeat. [00000006] [08/02/2014 16:38:43.62] [INFO] Role cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 is reporting state NotReady with sub-status Unresponsive and details Role cannot be reached by the host system.. [00000010] [08/02/2014 16:38:44.63] [INFO] Role cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 has current state Unresponsive, desired state Started, and goal state execution status StartSucceeded. [00000004] [08/02/2014 16:38:48.66] [HEART] WindowsAzureGuestAgent Heartbeat.
and
[00002052:00004500, 2014/08/02, 16:38:40.947, WARN ] HealthTimer for cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 expired. [00002052:00004500, 2014/08/02, 16:38:40.947, IMPRT] State of cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 changed from RoleStateStarted to RoleStateUnresponsive. [00002052:00007628, 2014/08/02, 19:30:51.309, IMPRT] State of cc94138f6d274dfd831683f4208b9f90.Ascend.Host.WebRole_IN_0 changed from RoleStateUnresponsive to RoleStateStarting.
this is from the two AppAgent logs.
I have found out that if I remote desktop into the machines and manually do
net stop w3svc net start w3svc
then i can get it to restart the health checks and report ready again. Otherwise I have to redeploy to get it to reset the state. Is there anyhting I can do to find out why it stop?