*UPDATE DOCKER Pi-HOLE*

    Zitat von raulfg3

    how do you update your docker?

    With Watchtower and scheduled job with this script :

    Code
    docker images -q --filter "dangling=true" | xargs -n1 -r docker rmi


    When i go in the pihole-FTL.log i have :


    which turns in a loop and this in pihole.log :



    Code
    Jan 31 13:29:55 dnsmasq[17404]: setting capabilities failed: Operation not permitted
Jan 31 13:29:55 dnsmasq[17404]: FAILED to start up
Jan 31 13:29:56 dnsmasq[17417]: setting capabilities failed: Operation not permitted
Jan 31 13:29:56 dnsmasq[17417]: FAILED to start up
Jan 31 13:29:57 dnsmasq[17430]: setting capabilities failed: Operation not permitted
Jan 31 13:29:57 dnsmasq[17430]: FAILED to start up

    AMD Ryzen 5 2400G on Asus TUF B450M-PLUS - 16Gb RAM - 3 * 3To RAID5 on LSI Megaraid SAS 9260-8i and 3 SSD in Fractal Design Node 804 Black
    OS: OMV 6.3.2-1 (Shaitan)

    2 Mal editiert, zuletzt von Methy ()

    sorry, not enough knowledge to help.


    try to update FTL first time from pi-hole webGUI, I read somewhere that sometimes is useful when updates refuses to start FTL service.



    googling a bit i find this:


    https://discourse.pi-hole.net/…fused-is-ftl-running/7962


    https://discourse.pi-hole.net/t/pihole-ftl-wont-start/10328

    • Offizieller Beitrag

    I think you'll find that the latest release is the cause of the problem have a look here


    What you might be able to do is to modify the container with the changed settings using references by @raulfg3 on page 3 and from the updated Pi-Hole guide


    BTW you don't have to run that script to clean up Docker images simply adding --cleanup within the watchtower container under command will do the same

    • Offizieller Beitrag
    Zitat von Methy

    With Watchtower and scheduled job with this script :

    It is for this exact reason that I believe using Watchtower is a bad idea. (My opinion.) One can go from a working Docker to a dead Docker, "automatically". Moreover, users may not know that their setup doesn't work, until network Clients can't resolve hosts. With a bit of hands on, every 4 to 6 months, the update issue can be dealt with, without "surprises".
    ____________________________________________________________


    If using this How-To with the latest Docker image :
    The only additional parameter that's absolutely required is --cap-add=NET_ADMIN , in the Docker's extra arguments line.This is, currently, in the Pi-hole How-To.


    --dns=127.0.01 is required only when using pi-hole as a DCHP server and local resolution of local hostnames is required/desired. Outside of those conditions, this is optional.


    The second DNS, as an extra argument, doesn't make sense to me. As the Docker author explains it, this is a backup DNS entry in the event that Pi-hole's DNSMasq "malfunctions". As I understand it, this would be a (temporary ? ) pi-hole bypass. If security is a concern, setting a second DNS address, VIA the extra arguments, may not be a good idea. (If configured, the IP address of the host where unbound is installed would be an exception,)

    Thank you for your help I will look at this :)

    AMD Ryzen 5 2400G on Asus TUF B450M-PLUS - 16Gb RAM - 3 * 3To RAID5 on LSI Megaraid SAS 9260-8i and 3 SSD in Fractal Design Node 804 Black
    OS: OMV 6.3.2-1 (Shaitan)

    oh my god I found the solution by looking on the site: D
    https://hub.docker.com/r/pihole/pihole


    I had not seen that the prerequisites there was a change:


    --cap-add = NET_ADMIN


    and it is clearly marked ...:

    Code
    --cap-add = NET_ADMIN Required FTL DNS will fail to start without this setting


    Thank you anyway :) and sorry for the loss of time :(

    AMD Ryzen 5 2400G on Asus TUF B450M-PLUS - 16Gb RAM - 3 * 3To RAID5 on LSI Megaraid SAS 9260-8i and 3 SSD in Fractal Design Node 804 Black
    OS: OMV 6.3.2-1 (Shaitan)

    Einmal editiert, zuletzt von Methy ()

    i am hoping someone can help me. Been fighting with this for 3 days now. here is my environment:


    OMV 4 is running on a VM inside a VMWare ESXi 6.5
    using docker plugin
    Pihole is 4.2.2


    i have followed the videos from Technodadlife:

    Externer Inhalt www.youtube.com
    Inhalte von externen Seiten werden ohne Ihre Zustimmung nicht automatisch geladen und angezeigt.
    Durch die Aktivierung der externen Inhalte erklären Sie sich damit einverstanden, dass personenbezogene Daten an Drittplattformen übermittelt werden. Mehr Informationen dazu haben wir in unserer Datenschutzerklärung zur Verfügung gestellt.


    and i followed the guide from these forums:
    [How To] Install Pi-Hole in Docker: Update 01/18/19 - Adding Unbound, a Recursive DNS Server.


    i have followed each guide at least 3 times each some times more. each time i re-start the build, i delete the container, the network i created, and empty the folder contents it creates so i start from scratch each time


    everytime i am done and the save the settings i wait for the container to build and each time it never fully startes and gets into a loop where it will be "starting" then go to "unhealthy" and then exit and restart the process.


    the admin page will never load and i always see in the logs "DNS Resolution is currently unavailable


    according to the 4.2 image EV "ServerIP" is no longer needed HOWEVER, i have tried builds with it and without it and neither works


    i have also tried the following extra arguments and none have helped:
    -p 53:53/tcp -p 53:53/udp -p 67:67/udp -p 80:80/tcp -p 443:443/tcp
    -p 53:53/tcp -p 53:53/udp -p 67:67/udp -p 80:80/tcp -p 443:443/tcp --cap-add=NET_ADMIN --dns=127.0.0.1
    -p 53:53/tcp -p 53:53/udp -p 67:67/udp -p 80:80/tcp -p 443:443/tcp --dns=127.0.0.1 --cap-add=NET_ADMIN



    i REALLY REALLY have no idea what is wrong or why this wont work. i really hope someone can assist me


    thanks in advance

    • Offizieller Beitrag
    Zitat von Menz

    i REALLY REALLY have no idea what is wrong or why this wont work. i really hope someone can assist me

    Ok if I have read the 4.2.2 correctly you no longer need to add the Server IP to Environment Variables, the --cap-add=NET_ADMIN is no longer required unless using Pi-Holes DHCP, but it is says nothing about --dns=127.0.0.1.


    If the above is correct the extra arguments would be;


    -p 53:53/tcp -p 53:53/udp -p 67:67/udp -p 80:80/tcp -p 443:443/tcp --dns=127.0.01


    but according to @flmaxey post above the dns option is only required to resolve local host name resolution, perhaps there has been a change....again!!


    But what I do to resolve this, once logged in to Pi-Hole, is go to settings, dns, scroll down to Advanced DNS settings, check the Use Conditional Forwarding and enter the IP of my DHCP server and the Local domain name and that resolves IP to name resolution for local hosts.


    EDIT: another option as suggested by @raulfg3 is to add DNS1 127.0.0.1 and DNS2 1.1.1.1 to the EV rather than add it to extra arguments.

    • Offizieller Beitrag

    The config posted was real world tested (not a VM) but they might have changed the Docker, yet again. (The pi-hole folks update a lot.)
    I'll run another config with the latest image to see if a change is needed.

    • Offizieller Beitrag
    Zitat von flmaxey

    I'll run another config with the latest image to see if a change is needed.

    If there is a change needed to the howto it might be prudent to add a warning regarding Watchtower, there is an option within Watchtower to not update certain containers and Pi-Hole would be one of those.

    Zitat von geaves

    But what I do to resolve this, once logged in to Pi-Hole, is go to settings, dns, scroll down to Advanced DNS settings, check the Use Conditional Forwarding and enter the IP of my DHCP server and the Local domain name and that resolves IP to name resolution for local hosts.

    the web page never opens so i can t do this. i'll look at some of your other suggestions as well

    Zitat von flmaxey

    Even if all is working correctly, it still takes 3 minutes or so, before the the admin page will come up.
    Give me a bit to take a look at it.

    believe me, i am not being argumentive in any way. i understand it takes a bit for the web page to come up but i can tell you on my system its like clockwork... the container says "starting", then "starting health status" then "starting unhealthy" and then the container restarts itself and the process takes 4 minutes

    • Offizieller Beitrag

    Ok,, there is a new image. I pulled it, image ID e0869bf1a536, installed it in OM4, on real hardware and tested it with a network client. It was configured, exactly, by the "How-To" , line-by-line, and it works fine.
    ___________________________________________________________________


    But, I noticed you're running OMV in VM, in ESXi. Did you see @ryecoaaron 's note about setting your VM's virtual switch to " promiscuous mode"?
    If ryecoaaron made a special mention of it, it's safe to say that it's required.

    Zitat von flmaxey

    But, I noticed you're running OMV in VM, in ESXi. Did you see @ryecoaaron 's note about setting your VM's virtual switch to " promiscuous mode"?


    If ryecoaaron made a special mention of it, it's safe to say that it's required.

    no i did not notice that at all. i just made that change. prior to making that change i again tried building the container and putting the dns servers in as Environment Variables. although i noticed the logs looked different it still did not keep the container up longer than 4 minutes.


    now that i made this ESXi change... i wonder which way i should try to install the container???

    • Offizieller Beitrag
    Zitat von Menz

    i wonder which way i should try to install the container???

    Exactly , line for line, as it laid out in the How-To.


    The VM switch setting is a virtual networking consideration, of ESXi. (But, since I don't have it, I couldn't tell you the exact reasons why it's needed.)
    Don't forget to check that there are no spaces before the slashes "/" in bind points and mounts. While minor, that's a show stopper.

    Zitat von flmaxey

    Exactly , line for line, as it laid out in the How-To.
    The VM switch setting is a virtual networking consideration, of ESXi. (But, since I don't have it, I couldn't tell you the exact reasons why it's needed.)
    Don't forget to check that the are no spaces before the slashes "/" in bind points and mounts. While minor, that's a show stopper.

    WOOO-HOOO!!!!! i changed the ESXi setting and i followed the guide to the letter, i even included the EV "ServerIP" like in your guide even though the docker says its not needed anymore. and lo and behold the web page finally loads!!!


    thanks for everyone's patience with me, it is much appreciated


    now just need to figure out how to cinfigure it. it appears my routers DHCP settings are on or off no where do i tell it what DNS server settings to hand out... hmmmm but thats not anyone on here's issue but mine

    • Offizieller Beitrag

    Well, your DHCP server (usually your router) will usually set itself (the routers IP address) as the DNS.


    Find all entries for "DNS" in your router such as "static DNS" or "local DNS and set them to PI-hole. (Do not set a secondary or backup DNS.)
    If you want pi-hole as your DHCP server that's another matter. It works, but I have yet to fully test it.
    _______


    Otherwise, glad you're set. :)

    using Pi-hole as my DHCP server seems to be working for now. lets see how it goes.....



    one other (probably dumb) question.... if pihole is working, do i still need adblocker addons on my browser?

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!

Benutzerkonto erstellen Anmelden