This is a new install on Raspberry Pi . all Os and kernel updates applied before running the Curl Install script.
I have installed Pihole before but had to reuse the Pi for another task.
So running the vanilla install, Pihole -r or Pihole -up always gives the FTL GitHub errors . I have disable the Pihole whilst performing the Updates from the GUI.. Option DISABLE and take a custom value of 15 mins.
Try the updates and the same failure message again. I have tried a sudo reboot just as a precaution.
The GUI reports the DNS and FTL not running. Low cpu and memory usage.
I will be patient and give another try tomorrow in case some links are outdate and require to be refreshed.
@kkorkky Sorry to hear that this solution isn't working for you. I wasn't getting the the 404 URL issue either but the exact same one as yours.
Hopefully you can manage to get FTL running next try. Just if it can be of any help, here's exactly how I proceeded:
Do the initial Pi-Hole install and wait until it finishes (while getting the GitHub error)
Access the web GUI and login to it
Disable the blocking permanently by clicking the Disable button then reload the web GUI to make sure it shows Pi-Hole as being offline
Go back to the terminal, type pihole -r and choose to keep the configuration
Wait until the repair finishes
If everything went well, you should see no error at the end of the repair. It indeed seems like that the blocking is interfering with the install/update.
I think I have a native DNS issue. DNS was tested and working before the Pihole install. Now after performing an uninstall of Pihole, Sudo Reboot I have no DNS resolution of BBc.co.uk . Even changed the DNS server to be 1.1.1.1 and no luck.
Alright, so I tried doing the solution process once again just to be sure and it seems like that Pi-Hole doesn't even wanna disable the blocking anymore, wether I do it from the GUI or from CLI. I keep getting that same error on doing pihole -r.
@anon55913113 I get the exact same error as @kkorkky in the first post of this thread:
Error: Unable to get latest release location from GitHub
Doing a nslookup on any domain gives me: connection timed out; no servers could be reached
I believe I'm gonna give up on getting a Pi-Hole running on my network for now and wait until a fix is released so it works on initial install. Thanks anyway!
Decided to give it a try just for the sake of it and yep, it's working with the /etc/resolv.conf fix.
Thanks a lot @deHakkelaar and @anon55913113!
Now that the FTLDNS is working, I'm assuming I just need to run sudo sed -i 's/8.8.8.8/127.0.0.1/' /etc/resolv.conf to get back to using the local DNS since it's now up and running.
I can't help but seeing a funny side to this issue, Pi-Hole not being able to insall the DNS server used to be able to install that same DNS server. Anyway, hope a fix will be released soon for new installs.
Thanks again!
Edit: Oh right. Dumb me. I didn't remember the file would get overwritten.
If you have run pihole -r, it did that for you already:
pi@noads:~ $ cat /etc/resolv.conf
# Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8)
# DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
nameserver 127.0.0.1
I believe nameserver in resolv.conf gets changed too quickly into 127.0.0.1 without checking if a DNS daemon is actually listening on the localhost IP (pihole-FTL failed to start for example).
Consequently, all further operations that depend on name resolution will fail.
Quite a rough start for v4.
But oh well, the sooner issues get discovered, the sooner they'll be fixed (hopefully).
I believe @deHakkelaar's message should probably be marked as the solution so any user experiencing that issue can actually find it quickly and know it's working.
This is indeed a problem. The installer has grown old over time as all of us (core team and external contributions) kept to blow it up to even more and more functionality. It consists of several thousand lines of bash code and we stretch out its flexibility. A rewrite of the installer is in the works which will also address some other things more elegantly, but this is still in an early alpha phase and not yet really ready for anything. However, there is light at the end of the tunnel
For some reason, even while using the fix from @deHakkelaar, I keep getting that same error over and over.
Looks like luck's not with me today, is there a specific delay to wait for between the commands or..?
So far I've been getting two "working-only-once" fixes.
@anon55913113 By running the netstat -tulpen : grep :53 command, I only see a few lines talking about services like Lighttpd, SSHd, Avahi and DHCPd, but no mention of any DNS running on port 53.
I even tried running pihole disable && sudo sed -i 's/127.0.0.1/8.8.8.8/' /etc/resolv.conf && pihole -r so all the commands can run without delay and nope, I still have the same issue, sadly.
Edit: It looks like I just cannot disable Pi-Hole by any means, wether I try running pihole disable, sudo pihole disable or even try disabling it through the web GUI.
Alright, so just to be sure, I tried seeing what was in my resolv.conf file after trying the repair and here's what I found:
pi@pi-hole:~ $ cat /etc/resolv.conf
# Generated by dhcpcd
# /etc/resolv.conf.head can replace this line
# /etc/resolv.conf.tail can replace this line
Seems like that's where the problem comes from. Also, I do believe it was the right command at first, @anon55913113, the one you gave in your last reply doesn't return anything in the terminal.
Edit: Looks like the solution is simply to do a sudo nano /etc/resolv.conf, add nameserver 8.8.8.8 at the end of the file and then simply run pihole -r.
I didn't remember the file would get overwritten.
