I need to show theis error message when local server is not running. I need to know which message is delivering the right information.
The answer depends on goals of the message. If your goal is just to inform user about server status, the first message is appropriate, while, if your goal is to request user to start the server, the second message is appropriate. However, in my opinion, a message that combines both goals would be the best (unless you have strict message length limitations). The combined message is optimal, since it informs the user about both the situation (status) and how to fix this issue (request).
It also might be beneficial (if enough space is available) to add to the message a full command for starting the server, just to help user (obviously, administrator) save time by avoiding extra lookup of command parameters in documentation.
Server is unavailable
The requested server is unavailable. Please contact your support and describe your issue.
Often the application requesting the server doesn't know why it doesn't get a reply. The server may be shut down, offline, services on server malfunctioning or permissions have changed. I think it's better to tell the user that the server can't be reached instead of making possibly false assumptions.
Also I think that not every user should have server access. Only admins should have the option to start a server.
The universal rule of thumb for an error message is to convey
- What went wrong ("Service not reachable")
- Why it went wrong ("Process name X not running"; "timeout on URL"; "port blocked by firewall on URL"; "port closed on URL"; ...)
- What the user can do about it (even if just "problem already reported to Admin and will be addressed. Please try again later." "in shell run 'service start blah'")
So this 'universal' rule does seem to fit your situation, and also address points flagged by other answers.