mrforsythexter

So, I have been having a strange issue. Every other day or two, I start getting messages from VM's and Docker Containers, They are full, or they stop. No updates can be installed for the dockers containers. Because I just didn't have a clue, I restarted the machine from the web interface, it comes back up, but the array is offline, and the cache disks are missing (please remember they are showing as fine before the reboot). The cache drives look fine, reporting 381GB Free. So, I shutdown the server from the web interface and turn it back on via the button. Everything comes up and is fine for days, then it happens again. It's like the controller for the cache drives fails or something, but unraid is unable to tell this has happened. ls /mnt/cache/ /bin/ls: reading directory '/mnt/cache/': Input/output error The drives are still listed via `lsblk` I can't stop the array (via web interface), the other disks (array)(Spinning rust) are accessible and writeable on the same sata controller. Any advice or direction would be appreciated. tower-diagnostics-20221019-1921.zip

Hmm this is interesting, thanks for the example. the reason i don't normally create a custom network, is that I still want it accessible from the main hosts IP (so the mariadb accessible by other machines on the network) I guess my next next is to test it in this configuration. Thank you! I will feedback / mark solved once tested (maybe a couple days)

Ken-jl, are you saying the docker resolver 127.0.0.11 only works with containers not in bridge mode? I have a proftpd container, configured for mysql authenication. I have a mariadb container, called mariadb. From the proftpd container, I was expecting to be able to dig @127.0.0.11 for mariadb, and get a response back like 172.15.0.3 for example. Is this not how it should work?

I think the title pretty much says it all really. I can ping 127.0.0.11 however if I dig like dig @127.0.0.11 mariadb I get nothing back, it just times out trying to connect to the DNS server. I should mention that all external DNS resolution works and is passed though to my router (when not trying to resolve a local container name directly). This seems like a basic part of Docker, but I am unsure where to look next. At this stage, I am not even sure how the image is being given my hosts DNS, the resolve.conf is empty. Hoping someone can point me in the right direction.

Thanks I have moved to another ombi docker container.

Thank you for coming back, it maybe fixed.. but best I can tell, its not a stable release, so I wont be just checking that, seems to risky to me. Is there a better place for feature requests?

@binhex that is correct, I am starting to think that a reverse proxy might be an easier idea. Not sure if its worth mentioning, but your container is running on the bridge network, not sure if I should have move it to "Custom: br0" and given it an ip on my network.

It seems to be having an issue starting... Getting Stable Version () % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed Unzipping... gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now chmod: cannot access '/app/Ombi/Ombi': No such file or directory s6-applyuidgid: fatal: unable to exec /app/Ombi/Ombi: No such file or directory /start.sh exited 111 any advice?

hi @binhex would have loved that to be the answer, I have checked its loaded using lsmod | grep iptable_mangle iptable_mangle 16384 2 ip_tables 24576 5 iptable_filter,iptable_nat,iptable_mangl and completed a reboot, however this is still not working for other containers which have been passed through yours. (added the port to yours, change other container to network none and added extra param --net=container:binhex-delugevpn) I can see that container is using your network (PIA), but can't then get to the webgui of this other container. any more Idea's?

This having so many posts has made it hard to find if anyone has answered this before. Great container @binhex, and great support from @SpaceInvaderOne I followed the guides to get this setup and passing other containers through this one, no problem, working great.. apart from external access. I used to be able to access the other containers externally. So lets say I have open port 3579 and it was going to a local container. That was working fine. When I route it through this container, I can access it locally using [IP]:[port], however if I try and access from outside my network.. nothing.. dead. I assume looking at the routes there is something wrong here. You appear to be adding a route that sends traffic from our network (LAN_NETWORK) back out over ETH0, other sources (apart from dockers network) are sent out over tun0. My guess (but I can't tell, as I can't see the package manager you are using to install tcpdump) is that traffic is getting to my destination container, but routing back out over the tun0 for some reason. My other guess is that privoxy is actually getting the request with the DNS on it, and not understanding how to route it correctly as it doesn't know its local (DNS Edge device ip pinning) Are you able to help either discover the route that is needed to be added to the table, or other another idea, like using another container to proxy access to the hidden containers now? I know there are other options, like connecting to my home network via VPN, but, at least for me, I have a couple of these now (Work etc) and switch them all the time is not as convenient. Hoping someone with a little more networking knowledge than me can help.

Thank you, You are 100% right, something to do with kaspersky, as soon as I disabled, it worked.

I did have add block, however it made no difference, I debuged the open window code to :- openWindow('/webGui/scripts/tail_log&arg1='+log, 'Log for:'+name, 600, 900) The link is.. "/webGui/scripts/tail_log&arg1=libvirt/qemu/CentOS_Transcoder.log" However, this URI doesn't work in a new window either.

Here is my diagnostics if anyone can help. tower-diagnostics-20200704-1727.zip