Open charliik opened 2 years ago
up, anyone?
Are you using a vpn?
Are you using a vpn?
Hey!
Yes, using Private Internet Access.
Wireguard tunnel Port transfert Split tunneling activated Network card bound to both Qbittorrent instances
Have you had it active on both test machines? I would be wondering if it's that interacting with tcpip.sys
Absolutely, yes.
Did not think about that, I've made a post on PIA' subreddit to see if they've got any idea.
Okay. If you're seeding to private trackers only, you may wish to disable the VPN and see if the situation changes.
Unfortunately no, I can't try that
Okay. I've gone and downloaded poolman, here's my own system for reference. I have 250k torrents seeding across 10 qbittorrent instances. The unknown allocation at the top is just deduplication in windows server.
oh wow, thanks for putting your time in, is this Windows 11 too?
I might have to wait for PIA answer then
Windows Server 2022 which is based on 10 (build 10.0.20348 per msinfo32)
Alright ok, thanks
I believe this is due to some faulty driver introduced in Windows 11.
@glassez @Chocobo1 any thoughts? I've seen multiple tickets stating about a huge non paged pool memory allocation when qBt is running. These allocations are usually pointing towards some network related drivers. Reference https://github.com/qbittorrent/qBittorrent/issues/15560 https://github.com/qbittorrent/qBittorrent/issues/16775
Just thought I'd chime in again.
As someone who also has qBittorrent bound to a VPN, I faced the same issue in Windows 11. It was getting to the point where I'd have to restart my computer every few days just to have enough RAM. After some rudimentary searching on 'Acrl' (ALE connect redirect layer), it sounds like it relates to the 'Split Tunnelling' feature that a lot of VPNs have nowadays. Sure enough, disabling that resolved the issue.
Easiest step to reproduce for me was to track the KBytes column of Acrl in PoolMonX and restart qBittorrent over and over again. Not sure yet if this issue is specific to qBittorrent? Could be a Windows 11 bug? The VPN I use is not the same as OP. Changing the VPN protocol made no difference for me.
Just thought I'd chime in again.
As someone who also has qBittorrent bound to a VPN, I faced the same issue in Windows 11. It was getting to the point where I'd have to restart my computer every few days just to have enough RAM. After some rudimentary searching on 'Acrl' (ALE connect redirect layer), it sounds like it relates to the 'Split Tunnelling' feature that a lot of VPNs have nowadays. Sure enough, disabling that resolved the issue.
Easiest step to reproduce for me was to track the KBytes column of Acrl in PoolMonX and restart qBittorrent over and over again. Not sure yet if this issue is specific to qBittorrent? Could be a Windows 11 bug? The VPN I use is not the same as OP. Changing the VPN protocol made no difference for me.
"Acrl" tag indicates the issue is happening in the WFP redirect layer so there's two main options (in my opinion): a) Common VPN behavior (i.e. both using the same WFP code) is allocating but not releasing when registering redirects with a split tunnel active b) There's a Microsoft bug that's doing the above through no fault of the VPN
i'm also experiencing this issue with qbittorrent 4.5.0, PIA 3.3.1, and Windows 11 Pro 22H2 22621.1105 :(. only started occurring after i switched from using a SOCKS5 proxy to PIA's VPN with split tunnel on and qbit bound to the VPN adapter
Using poolmon to track tcpip.sys "ALE connect redirect layer" memory usage while using Private Internet Access (PIA) VPN product and qBittorrent consistenely reports this to continually increase, and never decrease. No combination of settings between qBittorrent and PIA regarding OpenVpn, Wireguard, tcp, udp, interface selection, etc appears to resolve the problem with the slow but constant increase. After several days, the system becomes unusable as the Non-paged pool increases to multiple gigabytes, requiring a reboot. (This has been going on for years with the same qBittorrent+PIA combo, even with Win 10) When tesing an alternate VPN service, Proton VPN, "ALE connect redirect layer" does not even appear in poolmon's list, and I can not see any other consistent increase in any tcpip.sys related sources. Comparing with another service, NordVPN, that offers split-tunnel, it does appear that this memory leak pertains specifically to when split-tunnel is enabled. I will know in about two days if I also still see a severe increase in the Non-paged pool memory.
Exact same problem with qBittorrent and ProtonVPN (Wireguard and split tunneling). Acrl (tcpip.sys) currently eats 20GB of my RAM after 11 days of uptime.
Just joining in with some more info. Windows 11 22H2, qBittorrent, ProtonVPN with Wireguard and split tunneling. 12GB to tcpip.sys after 7 days.
Still seeing this with lastest Win 11, qBittorrent, ProtonVPN, Wireguard and split tunneling. On my 32GB system, I jump up to 16GB within a day or two so ram usage is in the mid 90%. Restarting qBittorrent or Proton doesn't release all the page file, only rebooting does it.
Can confirm this is still an issue; I’ve only noticed this behavior in the past few months, but I’m experiencing all the same symptoms. Windows 11, ProtonVPN with split tunneling, qbittorrent bound to my VPN and left running, and a gradually increasing nonpaged memory pool that will eventually eat all available memory unless I reboot my computer, which is frustrating, since mine doubles as a server… I found this thread after googling the “acrl” tag in poolmon, which seems to be exhibiting a similar behavior on my system as others have shown here.
has this issue been brought to Microsoft? it seems to be a bug in windows 11 tcpip.sys.
has this issue been brought to Microsoft? it seems to be a bug in windows 11 tcpip.sys.
I believe that this is the correct route to go, as I've seen this same complaint made on other torrent clients and multiple VPN applications/providers.
Does anyone have an active windows feedback post we can upvote to help bring attention to microsoft? This is still an issue.
I did a quick search and didn't find any that were established, so here's my post https://aka.ms/AAqfgbj
Edit: I have reached out to Privateinternetaccess and helped them create a ticket with Microsoft and they are investigating it on their end. I recommend people with other VPN's with the same issue do the same thing. The only way to solve this is to bring eyes on it.
What's the situation on this with qBttorrent 4.6.5 & latest/fully updated Windows 11 23H2 build with up-to-date network drivers etc.?
I stopped having this problem about a month back. Had fully uninstalled qbittorrent and protonvpn. Not sure if it was an update to one of the tree factors, or the uninstall/reinstalls but I haven't had the issue since, despite multiple updates to all three. Still running split VPN.
What's the situation on this with qBttorrent 4.6.5 & latest/fully updated Windows 11 23H2 build with up-to-date network drivers etc.?
The issue remains for me. qBittorrent 4.6.5, Windows 11 23H2 Build 22631, all drivers up to date. The Acrl tag increases by ~40 KB every 10 seconds. Privateinternetaccess V3.5.7 using WireGuard protocol. Intel N100 processor
qBittorrent & operating system versions
qBittorrent: 4.4.3 x64 Operating system: Windows 11 Pro 21H2 (10.0.22000) x64
What is the problem?
I've 2 Qbittorrent instances running (one for arr and one for rss seedbox use) which create a non paged pool memory leak through the tcpip.sys (see screenshot), only when I'm using Windows 11.
Got the same problem a while ago on another machine (Ryzen 5 5600x), switched again to W10 and it disappeared.
Now that I've got a dedicated machine for this, with absolutely different hardware (i5 11400) , I tried again W11 and yeah, it still happens.
I've tried multiples ethernet drivers, including the lastest one, no luck whatsoever.
Steps to reproduce
Additional context
Log(s) & preferences file(s)
No response