Receiving the above error for just one user. Any ideas ?
This is after logging in and being sent to /pun/sys/dashboard.
Users error.log is empty, access.log contains some HTTP 499 responses, not sure what to make of that.
Are you running OOD behind an HA proxy or similar?
It is running behind an AWS Load Balancer.
Is it possible that the user is timing out somewhere ? They are running about 7 different jobs, do they need to load each time they log in for some reason ?
time_proxy="60060.406 - This is the shortest time I see in the logs, some are way longer, in the six digits.
Something is timing out. 499 is the http code for the client disconnecting. I’m guessing in this case the ‘client’ is OnDemand.
Not sure what the issue could be, but googling 499 aws load balancer has a lot of hits.
This seemed to work but I’m not sure why.
cd /var/run/ondemand-nginx/username/
mv passenger.pid /tmp
mv passenger.sock /tmp
/opt/ood/nginx_stage/sbin/nginx_stage nginx_clean --user username
We also received " Website is under heavy load (queue full)" for that user, no one else experienced that.