Anything I can do about connection drop-off?

I’ve been running a bridge since May 2018 with the following config:

BridgeRelay 1
ORPort 8101
ServerTransportPlugin obfs4 exec /usr/bin/obfs4proxy
ServerTransportListenAddr obfs4 0.0.0.0:9101
ExtORPort auto

According to metrics.torproject.org, it’s been running at about 15 connected clients with a sudden spike between about 2020-04 and 2020-09 (up to 60 for a while), then again 2021-12 and 2022-02 up to about 40, whereupon it sank down to about 10 per day and falling.

Is this latest collapse a Russian thing? I see it says it’s blocked in RU. Should I try tweaking the config? I can’t change the IP address, unfortunately.

1 Like

Hi @gilgongo, could you tell us what is your bridge distribution method?
You can check that by searching your fingerprint on Relay search.

2 Likes

It says it’s “https”

Perhaps your bridge was assigned to moat or other bridgeDB distributor back then, but after rdsys deployment on February, maybe it’s using a not-so popular distributor?

If you want to investigate more about it, you will need to search your bridge fingerprint on ‘bridge pool assignments’ data on Metrics portal and see if my hypothesis make sense:
https://metrics.torproject.org/collector/recent/bridge-pool-assignments/

You can also change your bridge distributor method and use a more popular method, like moat (but it won’t be distributed to users in Russia):

I also recommend checking this interesting graph about bridgeDB requests by distributor:
https://metrics.torproject.org/bridgedb-distributor.html

3 Likes

Ah - thanks! There does seem to be a bit of a correlation between the distribution method rate and the popularity of my bridge, although I think the RU blocking may have had more of an effect overall.

In general though, I suppose if I changed to moat that would reduce the pool of https bridges, so maybe I better not for diversity’s sake?

Well, it’s up to you, both options are good. Thank you for running a bridge! :slight_smile:

1 Like