Dec 18 18:19:02 turris snowflake-proxy[18473]: 2022/12/18 18:19:02 In the last 1h0m0s, there were 17 connections. Traffic Relayed ↑ 177 MB, ↓ 16 MB.
Dec 18 19:19:02 turris snowflake-proxy[18473]: 2022/12/18 19:19:02 In the last 1h0m0s, there were 17 connections. Traffic Relayed ↑ 49 MB, ↓ 4 MB.
Seems to work 
Btw, Turris OS 6.1.0 was released. It includes snowflake version 2.4.1.
Does /etc/init.d/snowflake-proxy get overwritten on an update? Just noticed, my settings are lost after the upgrade to Turris OS 6.1.0.
Yes, it will be overwritten by each update of snowflake-proxy.
Thanks!
Is there a way to avoid this?
from TOS 6.1.0 with version 2.4.1 it worked, that was probably the reason.
From the moment I replace -verbose with -capacity I have no more entries in the log, is this normal?
I hope you added -capacity 10 (or another number). You should see at least something when you restart or start up snowflake-proxy, something like this:
snowflake-proxy[18825]: 2022/12/22 17:05:16 Proxy starting
And after starting snowflake-proxy every hour a summary of the connections that went over the snowflake-proxy is logged.
Okay, then everything is correct. I just entered exactly this value (-capacity 10) and it is as you described.
After the jump to version 2.4.1 with TOS 6.1.0 it also works with the firewall, probably that was the reason.
Thanks a lot!
Thank you, these settings and starting snowflake proxy with the following settings
snowflake-proxy -verbose -ephemeral-ports-range 32768:60999 made my connection unrestricted and now I don’t get the following error anymore “Timed out waiting for client to open data channel.” I used to have it,
I hope opening these ports don’t make a security problem.
Go to Backup / Flash Firmware > Configuration and add on a new line /etc/init.d/snowflake-proxy. This will save that file when you export a backup so you can restore it after a system upgrade.
did you put it all in here?:
/etc/init.d/snowflake-proxy
Does that work here too?:
procd_set_param command “$PROG” -capacity 10 -ephemeral-ports-range 32768:60999
No I didn’t. I just put it in the crontab so that at restart, It gets restarted with these configs
Btw. I still get timeout messages but much less than before.
Do not choose the value for capacity that high, as the man page of snowflake-proxy says the default value for -capacity is 10. So choosing a higher value than 10 increases the number of connections.
SNOWFLAKE-PROXY(1)
I think the high load average might be caused by a memory problem which is already discussed on GitHub and there are already possible fixes and merge requests for further versions