Status Update
Comments
ca...@google.com <ca...@google.com> #2
The App Engine Flex environment depends on Compute Engine for the instances, the instances follow its
As well, in the same physical server there could be other VM’s from other users. App Engine will constantly be checking the health of the VMs and the physical servers in order to provide the best performance for all the allocated VMs in the server.
When App Engine detects that the App could have better performance in another server or that there could be a potential failure if the VMs continue in that physical server, then the VMs are migrated. In fact those migrations happen all the time and are not perceptible at all but some apps can be sensitive for this.
For now the recommendation is to set 3 instances in the service.
ed...@go-smart.com.mx <ed...@go-smart.com.mx> #3
I understand the recommendation and will evaluate it's cost and benefit.
Does App Engine Standard also follow this scenario?
Description
My App Engine service is having downtime when being automatically rebooted.
What you expected to happen:
Having 2 fixed active instances I would expect to have no downtime.
Detailed config params:
runtime: nodejs
api_version: '1.0'
env: flexible
threadsafe: true
env_variables:
dev: 'false'
manual_scaling:
instances: 2
network:
name: default
networkSubnetworkName: default
resources:
cpu: 1
memory_gb: 1
disk_size_gb: 10
liveness_check:
initial_delay_sec: 300
check_interval_sec: 30
timeout_sec: 4
failure_threshold: 4
success_threshold: 2
readiness_check:
check_interval_sec: 5
timeout_sec: 4
failure_threshold: 2
success_threshold: 2
app_start_timeout_sec: 300
Thanks.