Posts by olduser

Quote from tocoson

Some of the commands

Then I'm wrong. I was under the impression that that example would execute 'ping' irrespective of the VPN connection.

Quote from tocoson

Here's the solution I found for qbittorrent and gluetun:

Are you sure that is utilizing the VPN for the kill switch? I thought I read that the health check will be ran by the docker process globally and not as the client software, thus that ping will always pass if there's a internet connection on the host system (or at least available to the docker process).

Quote from monsen

SMART values with a uart ttl convertor still works over a terminal connectio

If whatever ever worked for SMART, I'm sure it still does as SMART is seemingly feature frozen. However I don't see why that's needed as the drive is already connected via serial.... is SMART a "protected" feature?

I dont know. Either Yahoo doesn't have the very or your postfix cant find it. There's this from StackOverflow, not sure if it helps https://stackoverflow.com/ques…ls-connection-established

Also, it looks like you're using Yahoo's e-mail service, so do you need any of this?

Well if it's extremely private you can tell people to use a magic number in their emails and reject any email without it. I think there's a option for this in postfix or more widely in some GNU mail util (i used it ~20 I years ago).

As long as both daemons use it. Keep in mind nobody else sending you mail will be under any verification unless they too have it. This is why people use certificate authorities like letsencrypt.

FWIW, there's a docker stack called mailu that has a tutorial on this ... mailu.io

Quote from monsen

Any suggestion on how i could get to the root cause?

It's not worth it as you'd re-implement what the GUI already does. OMV isn't stateless so if you decide to change state without OMV to observe, you have to change everything that matches the last state OMV observed. I suggest you create a little switcher-roo script in $HOME and use that. It's like gderf said, cloning the drive is the best way here.

Quote from chente

Wouldn't it be a little sad to have to solve it that way?

Yes it is. Disabling hyper-threading (HT) isn't a cure, but it might side step the disease.

I suggested disabling it because if there's a cooling problem then HT isn't helping and if something is buggy with the microcode disabling HT could help if there's no BIOS fix (if it's really a BIOS problem). I'm imagining a cooling problem where the system is waiting for a cooldown so it can spin the CPU backup. If it's a code problem, then you'd almost certainly have to disable HT (or get a BIOS fix). Either way, it helps debug.

Is disabling Hyper-Threading not a option?

I don't really need it I don't think, I was just curious about the limitation of such a thing.

Quote from KM0201

wetty plugin is only available for amd64...

I'm curious, what creates this dependance?

Quote from fahrer17

would this configuration be a reasonable choice?

yes

175usd: HDPLEX 500W GaN Passive AIO ATX Power Supply

145usd: HDPLEX 250W GaN Passive AIO ATX Power Supply

The 250 watt might power 8 HDDs and everything else and its physical size opens up options.

Nope, that's a high wattage CPU (65w TDP).

All search results I see are seemingly related to a filesystem, almost exclusively ZFS.

The message of "prev->next should be..." makes it seem like a linked-list was modified without notifying the kernel. I have no idea how you'd sanely debug that without some hints on what is causing the panic.

Quote from gderf

such as having to add in some IP tables stuff.

Remove it. Remove everything related to iptables or it will conflict with the config inside the container. Apparently some VPN providers will add a "kill switch" config or something similar to their configs. For example, in the mullvad configuration tool you can toggle a switch to add in this extra stuff, otherwise it's fine. If your "wg0.conf" config looks like the below, you're fine.

[Interface]
# Device: Giving Alpaca
PrivateKey = abc
Address = 10.64.230.52/32,fc00:bbbb:bbbb:bb01::1:e633/128
DNS = 10.64.0.1

[Peer]
PublicKey = abc
AllowedIPs = 0.0.0.0/0,::0/0
Endpoint = 146.70.116.130:51820

With OpenVPN, I couldn't get the prefixed <ca> section to work, so I still had to have the .crt present along with the conf

Docker hub is a centralized repository for docker images and not a gated interface to a specific protocol. Of course if you're talking about some SDK or API docker hub offers for whatever reason, that will not be of interest to the client.

Quote from Soma

I might sound naive but what's the pros and cons of using the non-vpn qbit if you only use private torrents?

It has been 10+ years ago but when I belong to iptorrents leakage was always a problem. I'm not sure how non-members found the seeders but when they did everyone without a VPN was, in theory, "at risk".

Quote from gderf

why don't you fork the original

Honestly this seems more like a maintenance feature than something that should be added to the container. qbittorrent has a API to communicate outside of the container, but there's nothing else in the container that can be queried. It *might* be cleaner to setup a small http server with python that has some endpoint to query for other data... for instance the result of a speedtest. Of course, it would depend on how much more you're going to add in the future. But of course again, there is no current way to get data out of this container.

Quote from BernH

In this case though a background daemon whould not see a problem and therefore would not restart the container.

What was your case? I had a similar issue as stated in post 23 in this thread.

Quote from BernH

Unless it was doing periodic connection and speed testing

Well yeh, that would be the point.

Does the python 3 environment in the container have the json module?

Just type:

>>python3

>>import json

>>exit()

to find out.