On 23/07/2010 5:46 PM, Jeff Jensen wrote:
Hi Ceki, thanks for the reply. Welcome.
Yes, I upgraded to 0.9.24 while investigating this problem. I must have captured the stack before that.
OK.
I wondered about a race condition too. To try mitigate that at SubethaSMTP start/shutdown, I added Thread.sleep(10000) calls after start and before shutdown. There was no change; is 10 seconds long enough to wait to see if a shutdown server is the problem?
Yes, 10 seconds is more than enough.
Any suggestion for setting the timeout value via Logback?
As I mentioned in my previous mail, we also use SubethaSMTP for testing purposes. See SMTPAppender_SubethaSMTPTest [1]. Could you copy and paste SMTPAppender_SubethaSMTPTest into your environment and see if it runs properly? If it runs OK, I suggest comparing the code in SMTPAppender_SubethaSMTPTest with your test code. [1] http://logback.qos.ch/xref-test/ch/qos/logback/classic/net/SMTPAppender_Sube... -- Ceki