A colleague set up the Couchbase cluster and incorreclty set the email alert hostname to 10.15.2.200.
This was changed last week to 10.15.1.200 (and is correctly reflected in the console). The ‘test email’ sends correctly.
However, yesterday two nodes fell over, and neither was able to send an email - the error in logs contains the OLD address:
Could not send email: {error,retries_exceeded,
{network_failure,“10.15.2.200”,
{error,ehostunreach}}}. Make sure that your email settings are correct.
Any ideas?