MySQL Bugs: #45984: if the mysqld repository goes down, the ui can appear hung for 50 seconds

Bug #45984	if the mysqld repository goes down, the ui can appear hung for 50 seconds
Submitted:	6 Jul 2009 22:09
Reporter:	Sloan Childers	Email Updates:
Status:	Verified	Impact on me:	None
Category:	MySQL Enterprise Monitor: Web	Severity:	S3 (Non-critical)
Version:		OS:	Any
Assigned to:	Assigned Account	CPU Architecture:	Any

Description:
If the internal MEM mysqld repository goes down the UI can appear hung for as long as 50 seconds (the default retry/timeout setting) before the user is shown a stack trace.  

This is different behavior than other releases where we would show the user an error message on the next page refresh.

How to repeat:
Stop the MEM mysqld process and refresh the dashboard main page.

Suggested fix:
It would be nice to display a popup as soon as this critical problem is detected telling the user that the internal repository is currently unavailable for the duration of the outage or until we could successfully reconnect.

One could argue that we require the internal mysqld repository for MEM to function at all.  But on the other hand we've added code to keep us running in the case the internal mysqld repository is unreachable.  

This has changed how quickly we can notify an end user what the real problem is.

Gary Whizin writes: 
Good idea, nobody can think of a good way to fix it (similar/same behavior as other web-based apps)...

Raw ideas: 
* Could use Spring's transactions instead of managing our own and set transaction timeouts
* Could set/clear flag in BaseAction ??

Sloan Childers writes: 
Let's talk about what we can do here to notify a user that the backend is down.