I'm backing up a specific folder in a VM (server 2016) with DPM (SC2016). Both have all the updates installed, latest patch versions.
I noticed that after the initial replica, the backups have been failing to continue over the next few days. I noticed that the agent shows "Agent disconnected", after attempting a backup.
To troubleshoot agent connectivity, I started with simple "pings". I noticed that when attempting backups, the ping latency increased over time and finally stops responding.
Reply from 172.20.1.23: bytes=32 time<1ms TTL=128 Reply from
172.20.1.23: bytes=32 time=2109ms TTL=128 Reply from 172.20.1.23: bytes=32 time=3436ms TTL=128 Reply from 172.20.1.23: bytes=32 time=668ms TTL=128 Reply from 172.20.1.23: bytes=32 time=78ms TTL=128 Reply from 172.20.1.23: bytes=32 time=1719ms TTL=128 Reply from
172.20.1.23: bytes=32 time=3357ms TTL=128 Request timed out. Request timed out.
I left pings running overnight, with DPM backups disabled and the pings worked fine with 0% packet loss.
After this, I started running wireshark to see what's going on on a packet level. What I see is...
1) The source (server being backed up) shows "destination unreachable (protocol unreachable)" (type 3 code 2) message. According to RFC1122, this is related to "designated transfer protocol not supported". How the same protocol is supported when backups are not running is a huge mystery.
2) Even if the destination sends occasional replies, I get "request timed out" on the command line as a response to pings.
Additional info:
1) DPM server backs up other servers just fine.
I've tried:
1) Enabling/disabling adapter. 2) rebooting server (duh) 3) assigning different IP/Macs to the server being backed up. 4) removing/adding new network adapter to the server being backed up. 5) sfc /scannow to fix OS related issues
So, this is one of the weirdest network issues I've came across lately. Any idea what this may be caused by?
I'm considering rebuilding the entire server and removing adding the DPM agent again next.
