user12345678

Greetings. This issue was previously reported: And subsequently fixed in 6.11.1: However it's back now in 6.12.4. Looking closely at /etc/rc.d/rc.docker, there are two issues: Issue #1 - line 296: if [[ -n $MY_NETWORK && $MY_NETWORK != $MY_NETWORK ]]; then - The point of the if-statement is to skip processing MY_NETWORK from the inner loop if it is the same as MY_NETWORK from the outer loop. This situation would represent the case where we are looking at the default network that is defined in the XML, that will be processed like normal in the outer loop, hence the attempt to skip it in the inner loop. However, someone has done some 'clean up' and re-named the variables in this inner loop the same as the ones in the outer loop, specifically MY_NETWORK, which overwrites our only handle to the one in the outer loop. As-is, it just skips everything because MY_NETWORK is always equal to MY_NETWORK. In order to fix this, we'll need a new variable to hold the value of MY_NETWORK before this inner loop starts and we'll need to reference that in the if-statement, like so: OUTER_LOOP_NETWORK=$MY_NETWORK #save-off the value to use in the if-statement below for ROW in $USER_NETWORKS; do ROW=(${ROW/;/ }) MY_NETWORK=${ROW[0]} MY_IP=${ROW[1]/,/;} if [[ -n $MY_NETWORK && $MY_NETWORK != $OUTER_LOOP_NETWORK ]]; then LABEL=${MY_NETWORK//[0-9.]/} if [[ "br bond eth" =~ $LABEL && $LABEL != ${PORT:0:-1} ]]; then MY_NETWORK=${MY_NETWORK/$LABEL/${PORT:0:-1}} fi logger -t $(basename $0) "container $CONTAINER has an additional network that will be restored: $MY_NETWORK" NETRESTORE[$MY_NETWORK]="$CONTAINER,$MY_IP ${NETRESTORE[$MY_NETWORK]}" fi done Issue #2 - lines 297 - 300: LABEL=${MY_NETWORK//[0-9.]/} if [[ "br bond eth" =~ $LABEL && $LABEL != ${PORT:0:-1} ]]; then MY_NETWORK=${MY_NETWORK/$LABEL/${PORT:0:-1}} fi - I am guessing this has something to do with re-working network names when a user has applied the macvlan fix introduced in 6.12.4, maybe this tries to do an automatic 'translation' of sorts? It doesn't seem to 'translate' things very well, for me anyway it always just wants to change 'bond' to 'eth' so I get a bunch of 'eth2.10', 'eth2.20', etc. instead of what they should be - 'bond2.10' or 'bond2.20', etc. I have to completely comment this in order for the script to correctly restore my networks. It was obviously put here for a particular situation, but it seems to break other situations. My initial thought would be to just get rid of it entirely and let users know they will need to update their Post Arguments to correctly reference the proper network name(s), but I admit I am not aware of the exact, original reason this was put here so I may be missing something. Any thoughts? Thanks!!

Greetings. Any chance of including something like this in the 6.11 release? https://forums.unraid.net/bug-reports/stable-releases/6102-etcrcdrcdocker-ignores-additional-container-networks-and-cant-auto-start-affected-containers-r1983/ (I didn't look at any of the 6.11 RCs to see if this was in them, I just assumed not by lack of replies to my previous thread.) If not I can just keep doing what I'm doing, I just thought including this would help others who have ran into this & aren't keen on making these kinds of scripting changes, etc. in the core OS. Thanks!

Greetings. I posted about this in a few places hoping to gain some visibility, including here: So I'm trying a new 'bug report' now... The gist is that if the 'Post Arguments' field is used on a container (that is set to auto-start) to do something like '&& docker network connect...' with the intent of adding a second (or third, etc.) network to a container the rc script is unable to start the container on reboot, docker restart, etc. because it doesn't re-attach the network (after re-creating it) like it does for the main container network and since the container still references the old network id it fails with something like: unraid rc.docker: Plex-Media-Server: Error response from daemon: network ea31527bec520a923c5c8a466c0e265b775ba66262c7613d08684c786f5de5b4 not found unraid rc.docker: Error: failed to start containers: Plex-Media-Server My solution was to modify the rc.docker script in a hacky-way & add a loop to examine the container for additional networks and, if any are found, add them to the array that gets looped over later to re-attach. basically I added the following: # the loop above is based on the xml template which only defines one network - what about # containers where someone has done docker network connect <second network> <container>? # for those additional networks we need to cycle through them and add them to the array also ALL_CONTAINER_NETWORKS=$(docker container inspect $CONTAINER\ --format='{{range $key, $value := .NetworkSettings.Networks}}{{$key}}, {{if $value.IPAMConfig}} {{if $value.IPAMConfig.IPv4Address}}{{$value.IPAMConfig.IPv4Address}}{{end}} {{if $value.IPAMConfig.IPv6Address}}{{$value.IPAMConfig.IPv6Address}}{{end}} {{end}} |{{end}}'\ ) # an unfortunate side-effect of spreading the command above across multiple lines (for readability) is those newlines # sometimes end up in the final string, so take this opportunity to remove extra spaces and newlines from the result ALL_CONTAINER_NETWORKS=${ALL_CONTAINER_NETWORKS//[ $'\n']/} for CN in ${ALL_CONTAINER_NETWORKS//|/ }; do AN_ADDITIONAL_CONTAINER_NETWORK=${CN%,*} AN_ADDITIONAL_CONTAINER_IP=${CN#*,} if [[ -n $AN_ADDITIONAL_CONTAINER_NETWORK ]] && [[ $AN_ADDITIONAL_CONTAINER_NETWORK != $THIS_NETWORK ]]; then echo "container $CONTAINER has an additional network that will be restored: $CN" | logger -t $(basename $0) NETRESTORE[$AN_ADDITIONAL_CONTAINER_NETWORK]="$THIS_ID,$AN_ADDITIONAL_CONTAINER_IP ${NETRESTORE[$AN_ADDITIONAL_CONTAINER_NETWORK]}" fi done Immediately after line 197 (in-between the 'fi' and the 'done'). This seems to work really well and covers all the oddball situations that would otherwise be missed by either a go script hack or user script hack or some combination of that, however I admit I may have missed something. Any chance some cleaned-up version of this could make its way into the next release? Thanks!! EDIT: I didn't attach diagnostics since I felt I explained things pretty well & have a solution, plus I'd have to un-do the solution in order to produce the diagnostics. However if they must be supplied let me know and I can attach. Thanks!