Failed to install on raspberry pi 4 about hass-addons HOT 13 CLOSED

@kusha interesting, there must be some inconsistency between alpine's libc and the containerisation (based on the libc functions resulting in cordump in other crashes).

I'm going to pull your changes in and give it test now then might push a new release

from hass-addons.

Hey mate, sorry the add-on is causing issues for you!

The install.sh script downloads tailscale_VER_arm.tgz for both armhf and armv7 so the URL is fine. I think the actual issue is the container not being able to resolve or connect to pkgs.tailscale.com (i.e. the "bad address" error at the end).

This seems similar to #11. Could you quickly confirm the following for me?

• What kind of Home Assistant installation are you running? (HassOS/hand built etc)
• Could you retry installing the addon one more time (ideally 24h after the last attempt) just to confirm its not an intermittent fault
• Can you wget/curl pkgs.tailscale.com from within any other Home Assistant containers that are running?
• Do you have IPv6 connectivity from the network your Home Assistant install is on?

There's a fork of this repo with some additional testing and logging @ https://github.com/tsujamin/tailscale-badaddr-test if you want to try that and report back.

from hass-addons.

Hi

What kind of Home Assistant installation are you running? (HassOS/hand built etc)
supervised hass io
Could you retry installing the addon one more time (ideally 24h after the last attempt) just to confirm its not an intermittent fault
this i have done did not help -> tried in gaps of weeks also with restart
Can you wget/curl pkgs.tailscale.com from within any other Home Assistant containers that are running?
from within any other home assistant container also it work
Do you have IPv6 connectivity from the network your Home Assistant install is on?
no there is no IPv6 connectivity from my ISP

Output of the other test repository

Step 1/18 : ARG BUILD_ARCH
Step 2/18 : FROM homeassistant/${BUILD_ARCH}-base:latest
---> 564b1f71f9e5
Step 3/18 : ARG BUILD_ARCH
---> Using cache
---> 2ee8490de1e9
Step 4/18 : ARG TAILSCALE_VERSION="1.8.3"
---> Running in 1dc86ab3a720
Removing intermediate container 1dc86ab3a720
---> b792b98b25d3
Step 5/18 : ENV LANG C.UTF-8
---> Running in 41daf634d8ce
Removing intermediate container 41daf634d8ce
---> fd154f41233b
Step 6/18 : COPY install.sh /
---> ba2063052a8d
Step 7/18 : RUN chmod a+x /install.sh
---> Running in 85d6080ab604
Removing intermediate container 85d6080ab604
---> e12e41730394
Step 8/18 : RUN /install.sh $BUILD_ARCH $TAILSCALE_VERSION
---> Running in b852ef673a4e
ARGS: armv7 1.8.5
Downloading tailscale 1.8.5 for armv7
ifconfig
eth0 Link encap:Ethernet HWaddr 03:43:BC:11:00:03
inet addr:174.17.0.3 Bcast:174.17.255.255 Mask:255.255.0.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:6 errors:0 dropped:0 overruns:0 frame:0
TX packets:3 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:910 (910.0 B) TX bytes:306 (306.0 B)
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:65536 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
dns
time.c:126: Operation not permitted
timer.c:617: fatal error: RUNTIME_CHECK(isc_time_now((&now)) == 0) failed

Aborted (core dumped)

Removing intermediate container b852ef673a4e

Thanks
Sharath B K

from hass-addons.

Sorry I also forgot to ask what verison of HassOS/Supervisor/the underlying distro you're running. I'm just bringing my box up to latest to see if I can reproduce the issue.

Those second logs are super interesting because it shows the entire build process failing with a crash when I try and DNS resolve pkgs.tailscale.com.

time.c:126: Operation not permitted
timer.c:617: fatal error: RUNTIME_CHECK(isc_time_now((&now)) == 0) failed

Having a quick google around I find https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=132872 and visibilityspots/dockerfile-cloudflared#30 which seem related. In both cases there was an issue with the time being reported into the container subsequently breaking DNS... Can you double check the clock on your PI is being set correctly? Message #30 on the debian report may be useful, but honestly I'm grasping at straws here...

from hass-addons.

I did check time inside each container and also overall

However one thing i notice is that this supervisor container shows UTC

However i am able to do wget inside this container and down load the arm tgz. Does it still mean there is a problem with DNS ?

from hass-addons.

super interesting... I can't reproduce the issue but yeah its unlikely to be DNS specifically if it works in the other containers

I pushed another testing commit to https://github.com/tsujamin/tailscale-badaddr-test that hardcodes the domain name as a fallback. Want to give that a whirl and see if it builds and runs? If so I'll merge some fallback code into the main repo.

from hass-addons.

Hi Please find the log file with this test addon.

ipv4 shows failed however eth0 is

21-08-05 12:13:03 ERROR (SyncWorker_3) [supervisor.docker.addon] Can't build eed28774/armv7-addon-tailscale:1.8.5.1: The command '/bin/ash -o pipefail -c /install.sh $BUILD_ARCH $TAILSCALE_VERSION' retur ned a non-zero code: 1
21-08-05 12:13:03 ERROR (SyncWorker_3) [supervisor.docker.addon] Build log:
Step 1/18 : ARG BUILD_ARCH
Step 2/18 : FROM homeassistant/${BUILD_ARCH}-base:latest
---> 564b1f71f9e5
Step 3/18 : ARG BUILD_ARCH
---> Using cache
---> 2ee8490de1e9
Step 4/18 : ARG TAILSCALE_VERSION="1.8.3"
---> Using cache
---> b792b98b25d3
Step 5/18 : ENV LANG C.UTF-8
---> Using cache
---> fd154f41233b
Step 6/18 : COPY install.sh /
---> Using cache
---> c385603e2abf
Step 7/18 : RUN chmod a+x /install.sh
---> Using cache
---> de0e19bf7ee9
Step 8/18 : RUN /install.sh $BUILD_ARCH $TAILSCALE_VERSION
---> Running in 248f0f0fc262
ARGS: armv7 1.8.5
Downloading tailscale 1.8.5 for armv7
ifconfig
eth0 Link encap:Ethernet HWaddr 02:42:AC:11:00:03
inet addr:172.17.0.3 Bcast:172.17.255.255 Mask:255.255.0.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:6 errors:0 dropped:0 overruns:0 frame:0
TX packets:2 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:910 (910.0 B) TX bytes:196 (196.0 B)
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:65536 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
% Total % Received %
X
ferd Average Spee
d Time T
ime Time Curre
nt

Dload Upload Tot
al Spe
nt Left Spe
ed
0

0 0 0

• --:--:-- --:--:-- 0

• Could not resolve host: ipv4.tlund.se

• Closing connection 0

curl: (6) Could not resolve host: ipv4.
tlund.se

ipv4 failed
ipv6
% Total % Received %
Xferd Average Speed Time Time Time Current

        Dload  Upload   Total   Spent    Lef

t Speed
0 0 0 0 0 0
0
0

-:--:

:

:

:

:

0
*

Could not resolve host: ipv6.tlund.se

Closing connection 0

c
u
r
l
:

(
6
)

C
o
u
l
d

n
o
t

r
e
solve host: ipv6.tlund.se

ipv6 failed
dual
% Total
% Received % Xferd Averag
e Speed Time
Time Time Curren
t

                           Dload  Upload   Total   Spent    Left  Speed

0 0 0 0 0 0 0 0 --:--:-- --:--:--
--:--:-- 0* Could not resolve host: dual.tlund.se

• Closing connection 0

curl: (6) Could n
ot res
olve host: dual.tl
und.se

dual failed
% Total %
Received
% Xferd Average Speed Time Time Time Current
Dload U
pload Total Spent Left Speed
0 0
0 0 0 0 0 0 --:--:-- --:--:-- --
0 0 0 0 0
0 0 0 --:--:-- --:--:-- --:--:-- 0

curl: (28) Failed to connect to pkgs.tailscale.com po
rt 443: Operation timed out

tar: invalid magic

tar: short read

Removing intermediate container 248f0f0fc262

However this container has connectivity working fine

sudo docker exec -it hassio_supervisor bash
bash-5.1# ping www.google.com
PING www.google.com (216.58.196.68): 56 data bytes
64 bytes from 216.58.196.68: seq=0 ttl=118 time=44.961 ms
64 bytes from 216.58.196.68: seq=1 ttl=118 time=34.436 ms
64 bytes from 216.58.196.68: seq=2 ttl=118 time=34.111 ms
64 bytes from 216.58.196.68: seq=3 ttl=118 time=33.692 ms
64 bytes from 216.58.196.68: seq=4 ttl=118 time=35.981 ms
64 bytes from 216.58.196.68: seq=5 ttl=118 time=35.128 ms
64 bytes from 216.58.196.68: seq=6 ttl=118 time=34.352 ms
64 bytes from 216.58.196.68: seq=7 ttl=118 time=33.967 ms
^C
--- www.google.com ping statistics ---
8 packets transmitted, 8 packets received, 0% packet loss
round-trip min/avg/max = 33.692/35.828/44.961 ms

from hass-addons.

So strange... it's as if there's no IP connectivity at all in the built container. I wonder if its just during build time (i.e. setting up some CICD to prebuild containers would fix it) or its a problem with docker.

Could you check 2 more things (sorry!):

• sysctl net.ipv4.ip_forward from supervisor
• docker inspect de0e19bf7ee9 (or whatever the last layer before the one that fails to build is)
• check connectivity from within the last layer that didn't fail to build (docker run de0e19bf7ee9 curl google.com give or take)

from hass-addons.

Hi @tsujamin, I'm not sure if it is related to #11 or to this one, but I have a similar issue (supervised docker setup on amrv7 RPi 3b+). Sample logs:

Step 8/12 : RUN /install.sh $BUILD_ARCH $TAILSCALE_VERSION
 ---> Running in 893c375efb65
ARGS: armv7 1.12.3
Downloading tailscale 1.12.3 for armv7
Connecting to pkgs.tailscale.com (167.172.11.40:443)
ssl_client: pkgs.tailscale.com: certificate verification failed: format error in certificate's notBefore field
wget: error getting response: Connection reset by peer
tar: invalid magic
tar: short read

After adding --no-check-certificate attribute to wget:

Step 8/12 : RUN /install.sh $BUILD_ARCH $TAILSCALE_VERSION
 ---> Running in f93a3162f518
ARGS: armv7 1.12.3
Downloading tailscale 1.12.3 for armv7
Connecting to pkgs.tailscale.com (167.172.11.40:443)
writing to stdout
wget: clock_gettime(MONOTONIC) failed
tar: invalid magic
tar: short read

I forked the repo and switched the base image from Alpine to Debian: kusha@6f1f89f
The installation worked w/o any issues, so I'm pretty sure it is Alpine specific...

from hass-addons.

This fork @kusha mentioned worked for me aswell

from hass-addons.

1.12.3.2 released/rebased on debian (I cherry-picked your changes @kusha tyty) - does that alleviate the build issues?

from hass-addons.

@tsujamin I confirm that installation works from the upstream, ty

from hass-addons.

Awesome ty for writing it up and testing!

from hass-addons.