My server and mirror were working together happily until I shut them down and rebooted. Now the SurgeMail server and mirror won’t talk to each other. The error message on the status page is, “20:25 readline open No route to host after 0 seconds”.
From the mirror, I can get to the server admin interface in a web browser. I can also get to the mirror admin interface via localhost:7025
The mirror system can ping the server from the shell command line.
From the mirror, I can ssh into the host system.
From the server, I cannot get to the mirror admin interface in a web browser.
The server system can ping the mirror from the shell command line.
From the server, I can ssh into the mirror system.
From my usual workstation, I can bring up the server admin page in a web browser, but not the mirror admin page.
From my usual workstation, I can ssh into the mirror system.
From the mirror system, I can ssh into my usual workstation.
The traceroute command says there is a route:
$ traceroute mx.wilkes.com
traceroute to mx.wilkes.com (172.16.1.200), 30 hops max, 60 byte packets
1 mx (172.16.1.200) 0.432 ms !X 0.336 ms !X 0.286 ms !X
So it would appear that the mirror system itself has a route to the server system. Why does SurgeMail claim there is no route?
I have not changed any SurgeMail settings on either system. I have verified the mirror settings: g_mirror_mode, g_mirror_host (I use IP addresses instead of host names), and g_mirror_secret. They are all correct and unchanged from what they have been for months.
One of the advantages of being disorderly is that
I am constantly making interesting discoveries.