2.4.2.5 crash

2.4.2.5 is crashing when my sequence is aborting. I have some passing clouds coming through so the sequence goes into recovery. Twice, I manually stopped/aborted the recovery and thereafter I see an unhandled exception and crashes. Including logs but didn’t see anything obvious.

I do have email 1 configured (successfully) to text me errors and this is working fine, however in looking at the stack trace, the error appears to be a null reference exception somewhere related to sending emails. Including below (first line saying it was a null reference didn’t copy for whatever reason).

sg_logfile_20150812224046.zip (13.1 KB)

at System.Net.Mail.SmtpClient.set_Host(String value)
at SequenceGenerator.EmailEndpoint.a(EmailSettings A_0, String A_1, String A_2, Boolean A_3, Boolean A_4)
at SequenceGenerator.EmailEndpoint.e()
at SequenceGenerator.EmailEndpoint.f()
at System.Threading.ThreadHelper.ThreadStart_Context(Object state)
at System.Threading.ExecutionContext.RunInternal(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
at System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state, Boolean preserveSyncCtx)
at System.Threading.ExecutionContext.Run(ExecutionContext executionContext, ContextCallback callback, Object state)
at System.Threading.ThreadHelper.ThreadStart()

Not sure why this would happen if it sending any emails at all… it passes through that same code every time a message is sent.

Kind of hard to test, but gave it a quick try with just simulators since nothing is connected now and I couldn’t reproduce. But here’s one oddity: it had been working for me fine (getting texts via email, etc.) but I did disabled notifications after the second crash.

I went into the email notifications just now and tried to send a test message, and got an error: “Failure!: A recipient must be specified.” This was odd because there was a recipient specified, checked, verified, etc. I clicked update (for no reason, didn’t change anything) and then tried again, and it worked. So, maybe a fluke then, I’m not really sure.

I have found an issue where the endpoints were sharing some common resources that should not be allowed…

Great, thanks for the feature BTW, very useful. As an FYI, I tried going back a couple of versions so I wouldn’t worry about crashes with so many intermittent clouds. In doing so, I see ‘email’ and ‘email 1’ … ‘Email 3’ in the notification options. I didn’t try to make anything break however just wanted to mention this in case some cleanup/consistency checking is needed…

I reinstalled the latest and it cleaned it up fine…