[tor-relays] Huge drop in consensus and bandwith usage

[mattia]

Hi everyone! I've been running a relay
( 0D14A4498E7D1854696CC07548653EA4CD3CB2DA ) for more about 3 weeks
now.

Its lifecycle did not follow the path explained in the famous
"Lifecycle of a relay" blog post - I think because the same fingerprint
was previously used by the same server configured as a bridge.

My network capacity should be around 10Mbps (20Mbps nominal) in
download.

While consensus and bandwith usage have been growing for the first days
and then got stable (more than 200KB/s), since two days both values
have been dropping extremely
(https://globe.torproject.org/#/relay/0D14A4498E7D1854696CC07548653EA4CD3CB2DA).

The uptime is little because not knowing what to do I tried rebooting
the system - it did not help.

What could this be caused by? 
Thanks in advance


Mattia