We have a API built with node express, and some times a day the service get slowly.
The API is hosted in a VM with 8 CPUs and 27GB memory, dont have any other services hosted there. The VM is hosted in AZURE
We checked the server CPU and Memory is down, to be truth when this happen the CPU and Memory goes more down that usually are. We use PM2 to run 8 instances, that usually work and all 8 CPUs shows to be used, but again, when this happen just one look to be used with 100% and the others get with 0 use.
We have other VMs that hold Mysql and Mongo service, that also when this happen is with the resources free.
We have Newrelic to monitor our services, that shows a big "Truncated: Middleware: anonymous /" when this happened, that usually is not showed. It show up and goes down quicky, not in a incremented way.
https://www.dropbox.com/s/vrv5xrcm7g8peq6/Captura%20de%20tela%202017-04-13%2011.57.57.png?dl=0
We thought that could be a network issue between the VMs but when this happing the ping between all servers is less than 0.1 ms.
We have no more clue to go, how we can find out what is the "Truncated: Middleware: anonymous /sku", that new relic shows out ? this Middleware appear in all routes when the problem happen.
Thank you.
via André Fuhrman
No comments:
Post a Comment