Trafiek keeps failing - followed guide and stuck

sconnery

Respected Member
Original poster
Moderator
FreeLancer
Donor
Local time
6:21 PM
Jan 10, 2019
81
15
Sydney, Australia
PG Version
8.7.5
Server Type
Remote - Dedicated Server
hey team,

Tried to use another domain on my server to consolidate servers
keeps failing even though I followed it to a tee.
maxed out lets encrypt while trying to now need to wait any one had issues redeploying via cloudflare?
thanks
Steve
 

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
There is definitely something going on with Traefik. I have had this happen several times to me now on my test servers. The portainer check fails even though all of my settings are correct. The domain has been used successfully before and since these errors so there are no DNS record issues. Once this happens the only way I can possibly fix it is to rebuild the server and try again.
 

sconnery

Respected Member
Original poster
Moderator
FreeLancer
Donor
Local time
6:21 PM
Jan 10, 2019
81
15
Sydney, Australia
yes that is my exact issue, i destoryed traffik and redid the apis etc. same issue :/
 

spediegun

Full Member
Donor
Local time
2:21 AM
Mar 13, 2018
32
12
I’m not experienced at all but I suspect that if you try to set up traefik right after you make your cloudflare the portainer checks will keep failing resulting in some kind of temporary ban which will further delay traefik from setting up. Just like you I followed everything to a tee tried and tried and eventually gave up and came back one week later and changed nothing tried to deploy and to my surprise It worked. Maybe waiting and trying again will work
 

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
I’m not experienced at all but I suspect that if you try to set up traefik right after you make your cloudflare the portainer checks will keep failing resulting in some kind of temporary ban which will further delay traefik from setting up. Just like you I followed everything to a tee tried and tried and eventually gave up and came back one week later and changed nothing tried to deploy and to my surprise It worked. Maybe waiting and trying again will work
I use a long established domain on cloudflare that has not changed in months. I have a vps I use for testing deployment scenarios and what not. So same vps and same cloudflare domain will fail on some rebuilds and be successful on others. Once failed I have tried switching to a temporary duckdns domain but that fails also. So far only way to fix it is with a rebuild.
 
  • Like
Reactions: Admin9705

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
i built a new server and will test it out on it
FYI, I would say that traefik fails 50% of my deployments so if you do one build and it works you may have to do one or two more to see the issue.
 
  • Like
Reactions: DeadPool

Admin9705

Administrator
Project Manager
Local time
2:21 AM
Jan 17, 2018
4,760
1,809
usually it's a timing or DNS issue. With timing, I check back later and it works. It using the API from provider. Godaddy is pretty fast in the process.
 

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
Maybe but when I have run into this issue I have destroyed Traefik, deleted the /opt/appdata/traefik folder and then tried to redeploy traefik using a duckdns domain but the portainer check still fails.
 

Admin9705

Administrator
Project Manager
Local time
2:21 AM
Jan 17, 2018
4,760
1,809
i'll run this on my new server tonight and report back
 

sconnery

Respected Member
Original poster
Moderator
FreeLancer
Donor
Local time
6:21 PM
Jan 10, 2019
81
15
Sydney, Australia
its painful but alas still fun, i love troubleshooting this stuff, I might spin up a cheap vps to test redeployment.
thanks guys :)
 
  • Like
Reactions: DeadPool

DeadPool

Senior Member
Moderator
Local time
7:21 AM
May 2, 2018
162
35
hate to hijack guys but alas my (sorta/kinda/not really ;) q went unanswered elsewhere..
my google account disabling (edu account as per my OP) had further unintended consequences (SO stupid in hindsight) because i used the google account details as my login for duckdns.....and since account disabled i cannot now(obviously ....IN HINDSIGHT) log in to my duckdns to change the domain/release the domain so i can add/use it again.
duckdns devs (afaik) unreachable through advertised methods.
any ideas?


(once again soz for the 'jack)
 

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
Looks like they don't have any real support. Why are you using duckdns? Domains are dirt cheap and cloudflare is free and awesome. Can't you just make a new domain, duckdns or otherwise, and use that?
Post automatically merged:

Back on topic, this traefik issue is a problem. I cannot get it to deploy. The portainer check fails every time now. I had not deployed in over two weeks now. I tried to deploy again. First time I used a godaddy domain via cloudflare and the 2nd time with a duckdns domain. Has anyone been able to deploy traefik in the last few weeks?
 
Last edited:
  • Like
Reactions: DeadPool

sconnery

Respected Member
Original poster
Moderator
FreeLancer
Donor
Local time
6:21 PM
Jan 10, 2019
81
15
Sydney, Australia
i just rebuilt my server , took me a hour or so to rebuild and now all is running smoothly except transmission but that will be another thread soon :D
 

DeadPool

Senior Member
Moderator
Local time
7:21 AM
May 2, 2018
162
35
thanks for the reply despite the jack..
the truth is i had tried cf years ago but found it more complicated than duckdns.
but having just re-read the wiki it seems less so now(maybe i have a bit more knowledge now, who knows..).
I'll give it another whack.
i have used a new domain with a new duckdns account but my point is (IF I WANT TO USE DUCKDNS DOMAIN) that i cannot release the name i have used since the beginning of the project("USER already exists").
And duckdns is free (im a tight *unt) haha.
No big deal ill try cf as soon as 9.1 or 10 are out(rebuild from scratch).
another benefit of doing so (i had missed before) is using the Content delivery network.
having read a fair bit about cdns (conversations between people on this very forum) but never getting round to implementing it myself yet (im a lazy *unt) do you all swear by them and use them?
if so do they really make a useful difference?

dP
 

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
It's not hard to set up at all. I can help you out if you want. I would suggest getting a domain from google domains. Then create your CF account and add that domain to it. CF will tell you how to adjust your google domain. Once that is done you then create the appropriate DNS records on CF. That's it.

I live in Texas and use a hetzner server from Germany. Without CF my throughput to the hetzner box was horribly slow. With CF I get 10 times the throughput.
 

matt314159

Senior Member
Donor
Local time
1:21 AM
Nov 29, 2018
114
21
Back on topic, this traefik issue is a problem. I cannot get it to deploy. The portainer check fails every time now. I had not deployed in over two weeks now. I tried to deploy again. First time I used a godaddy domain via cloudflare and the 2nd time with a duckdns domain. Has anyone been able to deploy traefik in the last few weeks?
Did you get this sorted? I started having a Traefik warning when starting PlexGuide so I tried to redeploy and now I'm getting Portainer check failed no matter what I do. It was all running well this morning,.
 

UncleBuck

Senior Member
Staff
Local time
1:21 AM
Dec 21, 2018
199
55
Houston, TX
Thankfully it's working fine on my production plexguide server. I usually have a cheap vps for testing things and I get it to work there every once out of three or four attempts. Once it fails the only fix is to wipe your server and try from scratch. Keep doing this until you get it to work. Only problem with this method is you may hit the Lets Encrypt limit.
 

Xployt

Administrator
Moderator
Donor
Local time
7:21 AM
Sep 26, 2018
169
77

PlexFan

Respected Member
Staff
Local time
2:21 AM
Aug 8, 2019
53
8
The key for me for godaddy is to remove the txt file present at the domain if there prior which I noticed can be problematic
 

Xployt

Administrator
Moderator
Donor
Local time
7:21 AM
Sep 26, 2018
169
77
Okay sorry just to confirm. You've rebuilt and it's still not working?
 

sconnery

Respected Member
Original poster
Moderator
FreeLancer
Donor
Local time
6:21 PM
Jan 10, 2019
81
15
Sydney, Australia
no its working now a rebuild fixed it. but from your message looks like google auth was the culprite.
 
  • Like
Reactions: Xployt

matt314159

Senior Member
Donor
Local time
1:21 AM
Nov 29, 2018
114
21
rebuild as in blow PG completely away, or....? I'm in a real pickle here where Traefik won't deploy. It was working stable today until I did a server reboot, and PG gave me a warning message about everything being inaccessible. So I tried to redeploy Traefik and here we are.

I'm scared of having to start over from scratch with 3K+ movies, 500+ shows, etc, reimporting everything, setting it all back up...starting to panic a little.
 

matt314159

Senior Member
Donor
Local time
1:21 AM
Nov 29, 2018
114
21
It seems like people are saying oauth is the problem here, but what is the process for me to remove that so I can try to deploy traefik?

I'm kind of dead in the water here and starting to freak out a little. I really don't want to have to start from scratch.

EDIT - I need a drink. I redeployed pgshields and after clearing my browser cache and signing back in to oauth all my certificate errors and stuff are gone and plexguide seems happy. All my services are back up and running, fully accessible. I think I'm back in business but I don't understand what that was. So strange. I think you can safely disregard my posts in this thread, or I can delete them if staff prefer.
 
Last edited:
  • Like
Reactions: PlexFan
Assists Greatly with Development Costs

metrocard

Junior Member
Donor
Local time
1:21 AM
Jun 16, 2018
8
5
I'm running into issues setting up traefik with Cloudflare as well. I've tried it with two domains just to rule out Let's Encrypt rate limiting issues.

One thing I'm noticing is that the deployment is creating two _acme-challenge TXT records in the DNS. Is it possible it's confusing itself and calling the wrong one after both have been created?

I've also searched on https://crt.sh per the Let's Encrypt rate limiting page to see how many certificates have been issued, and both domains return zero results, so something seems to be off.
 

mackattack57

Full Member
Local time
2:21 AM
Apr 14, 2019
40
3
I am in the same boat here. I had an issue and now my server has been down for weeks. I am just getting some time off with the holiday to work on it. I am doing a clean build. I am going to do everything but OAuth and see if that works out. I always get the error after it so I believe that is the culprit as well.

I am using a Hetzner server with a Google and a GoDaddy domain to test both through Cloudflare.

I will let you guys know later today.
 

sconnery

Respected Member
Original poster
Moderator
FreeLancer
Donor
Local time
6:21 PM
Jan 10, 2019
81
15
Sydney, Australia
to revisit this my AWS vps is now having this issue. I tried reopening ports and removing google oauth and its still playing up.
 

mackattack57

Full Member
Local time
2:21 AM
Apr 14, 2019
40
3
I just did a build with PG Blitz Unencrypted. I had zero issues and deployed the build with no OAuth. However, once I started to build Plex and it accessed UnionFS mount, my server stopped responding and gave me the same issue. I then changed the mount to PG Move Unencrypted. It wouldn't deploy.

I feel the issue might be more related to using Drive. I had zero issues with it until my API calls started hitting the drive. I am not well versed enough to test this theory though
Post automatically merged:

I am going to try a build where I change everything to local and report back. That might be the best way to test that part of the theory. If it builds and communicates, I will add OAuth and report that too.

THAT DIDN'T WORK EITHER.

Going to attempt something different but this is getting really frustrating. I see it succeed at first deployment as first step of the process. It then gets fouled up somehow along the way that in not recoverable and causes me to wipe the server.
Post automatically merged:

SOOOO I got a working solution

I was able to get 100% up and running when I did a clean install of 8.6.5 FWIW
 
Last edited:

Create an account or login to comment

You must be a member in order to leave a comment

Create account

Create an account on our community. It's easy!

Log in

Already have an account? Log in here.

Similar threads


Staff online

Maintenance Donations

Recommend NewsGroups

      Up To a 58% Discount!

Trending