Restoring Mobile Platform Server Availability

Mobile platform server unavailability is characterized with errors during API user authentication and unavailability of administrator console in the browser.

To restore mobile platform server availability:

Check access to proxy server used in the corporate environment. To do this, check mobile platform server availability:

Execute request using the status API method via proxy server and directly to a mobile platform server:

curl "<http or https>://<IP address or DNS name of prepared server>/api/v1/status/"

Specify browser URL in the address bar depending on the protocol in use:

<http or https>://<IP address or DNS server name>/api/v1/status/

After executing one of the operations, a JSON string is obtained containing mobile platform server availability check result.

The successful result of mobile platform server availability check:

{"message":"Welcome to Foresight Mobile Platform.","time":"2019-03-19T12:35:53.390324"}

NOTE. A mobile platform server can be checked without authentication.

If after sending a request directly to a mobile platform server, the server returns a correct answer, and sending a request via proxy server returns a different answer, the corresponding proxy server is unavailable. Check incorrect work of the proxy server and solve the error. To quickly restore proxy server work without executing diagnostics:

Restore the proxy server virtual machine from backup and execute a request to check mobile platform server availability.
Make network routes active.
Check certificate validity.
Exclude possible activity of network protection system for some time.

If the proxy server works correctly but mobile platform server is unavailable:

Go to Step 2, if on installing Foresight Mobile Platform a fault-tolerant cluster is not deployed.
Check pods work on each cluster node and audit their logs if on installing Foresight Mobile Platform a fault-tolerant cluster is deployed based on Kubernetes, Deckhouse or OKD/OCP. The example of checking pods work is given for the OKD/OCP cluster.

Restart a mobile platform server. To do this, use the docker-compose utility and the docker-compose.standalone.yml configuration file:

Delete the running Docker components with data saving:

% docker-compose -f docker-compose.standalone.yml down

Restart the containers:

% docker-compose -f docker-compose.standalone.yml up -d

If centralized logging of mobile platform components is enabled, commands will differ:

Delete the running containers with data saving:

% docker-compose -f docker-compose.standalone.yml -f docker-compose.metrics.yml down

Start the containers:

% docker-compose -f docker-compose.standalone.yml -f docker-compose.metrics.yml up -d

Check status of Docker containers:

% docker ps -a

After executing the command, a table with information about containers is displayed: identifier, name, status, and so on. If container status in the STATUS column differs from Up, this container is not started. To start a container, execute the command:

% docker start <container identifier>

In the <container identifier> substitution use the container identifier specified in the CONTAINER ID column.

Repeat Step 1. A mobile platform server should return correct answer.
Open administrator console.

After executing the operations, mobile platform server availability is restored and administrator console is opened.

See. also:

System Error Monitoring