tjb_altf4

My setup has a direct connection between server and workstation, no switch. I'd probably look at the Unifi US‑16‑XG if I needed to connect more devices, as I'm already invested in that network ecosystem.

Try these DNS addresses: 10.0.0.242 10.0.0.243 https://www.privateinternetaccess.com/helpdesk/kb/articles/next-generation-dns-custom-configuration

Running Mellanox ConnectX-3 (dual sfp+ ethernet) on my Unraid server (not specifically that motherboard though), with the same on my win10 workstation, direct connection. Never had any issues, good performance, easy to setup etc.

This issue is still happening, logs show this on repeat (wireguard connection) 2020-11-09 10:59:03,623 DEBG 'watchdog-script' stdout output: 0 2020-11-09 11:00:38,903 DEBG 'start-script' stdout output: [info] Successfully assigned and bound incoming port '41561' 2020-11-09 11:15:39,838 DEBG 'start-script' stdout output: [info] Successfully assigned and bound incoming port '41561' 2020-11-09 11:29:30,654 DEBG 'watchdog-script' stdout output: 0 2020-11-09 11:29:34,772 DEBG 'watchdog-script' stdout output: [warn] Incoming port site 'https://portchecker.co/' failed to web scrape, marking as failed I can connect to UI and STRICT_PORT_FORWARD is enabled, so there is an established vpn connection (based on previous experience), but nothing actually connects. Restart the container and everything starts to connect again. Other than that its working quite well.

Not using ZFS, but this is a nice solution as it collects smart data so you can get historic data. Works for all drive devices, not just array/cache

Most trackers I'm on have a banned list that you cannot use, and an approved list that is supported/preferred, leaving a grey area for certain clients and/or versions of those clients. I would do some research on your tracker's forums etc to see if there is a grey area that allows use of these version of deluge.

Created to facilitate Host access to custom networks functionality you have enabled in Docker

Try the 6.9 beta, it has a newer kernel which I believe should have this driver as it landed in the kernel just after the intel/realtek 2.5G drivers (which are in the 6.9 beta).

Seeing an issue from time to time where vpn connection is active, but nothing downloads or uploads. I actually saw it in the last few days of using legacy + deluge 1.3, and assumed the old servers were cactus. The only fix is to restart the container then everything comes to life again. Is this something others are seeing? Maybe something happening in the PIA network?

What version of unraid are you on? 6.8 or later is required for Wireguard

Do you have privileged enabled in the template? Make sure you've set it as per Binhex's updated documentation for wireguard https://hub.docker.com/r/binhex/arch-delugevpn On my setup I haven't needed to use sysctl="net.ipv4.conf.all.src_valid_mark=1" arg, I'm not sure what conditions this is needed? @binhex

Agree with the feature request, but I don't think it will impact many users, yet. Rate limit kicks in at 100 pulls in a 6 hour period, however I could see a future where they would require all users to authenticate to pull images. If anyone hits this in the meantime, should be easy to setup a docker login script to run on array start.

The next step is usually where it uses the name server, so I would check what values you have set for NAME_SERVERS (DNS) For DNS, I use PIA's new (internal network), Google & Cloudflare 10.0.0.242,10.0.0.243,8.8.8.8,8.8.4.4,1.1.1.1,1.0.0.1

What time does your mover run? Mine runs at runs at 330am daily when the array is not active (faster moving) and the impact of the mover is unlikely to impact me.

Have you updated your delugevpn image to latest and used the next-gen OpenVPN files?

Logs are telling you exactly what is happening. PIA endpoint 'spain.privacy.network' is not in the list of endpoints that support port forwarding. Please consider switching to one of the endpoints shown e.g. spain.privateinternetaccess.com

"Yes" setting works with both next-gen openvpn and wireguard

Yeah not sure, like I said I still had a connection to the endpoint (Spain), but the endpoint was not connecting externally. That said, happy with getting this task done a week before PIA have said legacy is scheduled to be cut off.

OK seems that I needed to add the DNSWatch IPs from the newer template config to get everything working again. Finally migrated from my crusty v1 setup, v2 feels very snappy! Anyone have a good replacement for the seedtime plugin that works with v2 ? EDIT: Found a fork that has been updated for v2 compatibility, happy days vanossj/deluge-seedtime

Up and running with wireguard version inside of 5min on my testing instance, good work @binhex for making this so easy

Looks like Legacy is dead now, certainly the endpoint I'm connected to is no longer passing traffic (but allowing my connection to survive).

hotrodding i.e. cutting a hole and mounting a scoop (3d printed or the like) over the top is another thought I had to keep the existing cooler

Will be picking a U9 most likely, but I'll do a test fit first... it might just squeeze in with different fans

Managed to snag a good deal (free shipping) on a pair of Rosewill RSV-L4500s from Newegg, a rare opportunity for an Aussie customer, especially given these have been out of stock for most of the year. Looking forward to migrating my unraid server and workstation into a server rack in the coming weeks.

Same, guessing it's an issue with the upstream maintainer not updating package correctly