-
Controller Version5.12 Describe Your Issue or QuestionI upgraded as normal using portainer, pulling the new image. This time I get a password rejection, nothing obvious in the docker logs. Expected BehaviorAllow to login Steps to ReproduceUse password manager to attempt to login. How You're Launching the ContainerContainer LogsAdditional ContextMy emergency reset creds work fine, and I have config backups, so I can restore if needed, but adding this in if there's a simple fix, or an issue that will affect more people. |
Beta Was this translation helpful? Give feedback.
Replies: 11 comments 30 replies
-
|
It might be worth looking at the |
Beta Was this translation helpful? Give feedback.
-
|
So I took a full volume backup just to test, restored it with latest, password failure. Then tried the same backup (a fresh copy in case anything was changed) with the 5.9 version, all good. I wonder if there's something strange with my setup that causes this, because that sounds like something not on your end. |
Beta Was this translation helpful? Give feedback.
-
|
Same problem here, no logs present in my persisted volume folder. I don't have any recent backups other than a snapshot of the vm. Restored that which was v5.9.31 and all works well again. |
Beta Was this translation helpful? Give feedback.
-
|
I would be very curious if there is something unique about your passwords that some sort of data migration on the MongoDB backend is messing up; like special characters or something like non-ASCII characters. Are both of you running this on x86/amd64 or are either of them running on armv7l? |
Beta Was this translation helpful? Give feedback.
-
|
Mine is only accessible on my local network in specific vlan, cloud disabled. So I have it set to a crappy all letter password. I'm running on x86/64 - debian 12 vm using docker-compose. Haven't had this issue before, I use watchtower for automatic upgrades and then check back if something's broken (not the best approach but usually less time consuming on a whole then manually doing it) |
Beta Was this translation helpful? Give feedback.
-
|
x86-64, no special characters. I use bitwarden and don't have special characters selected to avoid any possible issues. I did create a backup user that I can attempt an upgrade at some point to see if I get locked out again. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks @Aevrin. So when you restore from a backup, do logins fail every time you upgrade again or is it random? Just curious how reproducible it is. |
Beta Was this translation helpful? Give feedback.
-
|
When I used the 5.12 version it would fail with every DB backup. When I tried with 5.9 with the same backup version no issues. I haven't tried the commit from 8 hours ago. |
Beta Was this translation helpful? Give feedback.
-
|
Sorry, I didn't ask this very clearly - So if you restore a 5.9 backup and then upgrade to 5.12.x, it always fails to login after the upgrade to 5.12.x? Or if you do the restore and upgrade multiple times, does it work sometimes and not others? |
Beta Was this translation helpful? Give feedback.
-
|
Tried doing an upgrade, got locked out again. Forgot to try the backup account. |
Beta Was this translation helpful? Give feedback.
-
|
Other thing I notice is that if I do this upgrade it locks down my internet by 3-5 minutes, and the same when I go back to a working version, and I don't recall this happening with other controller upgrades. |
Beta Was this translation helpful? Give feedback.
-
|
Also as a note, switching from .9 to .12 it has to readopt everything, and takes things down for a little, same when I downgrade back. |
Beta Was this translation helpful? Give feedback.
-
|
I see that there is a thread up on the TP-Link forums now: https://community.tp-link.com/en/business/forum/topic/623942 (thanks if one of you created it). I also see that someone on Reddit might have found a possible cause or at least the cause for one person. |
Beta Was this translation helpful? Give feedback.
-
|
I have the same issue. No weird characters in the password. |
Beta Was this translation helpful? Give feedback.
-
|
OK, so if someone who is comfortable doing some semi-advanced steps could try this out, let me know:
# stop the controller gracefully
docker stop -t 60 omada-controller
# create a docker network
docker network create -d bridge mongonet
# run mongodb manually as 508:508
docker run -itd --network mongonet --name mongodb -u 508:508 -v /mydata/omada-controller/data:/opt/tplink/EAPController/data mongo:3.6.8 mongod --dbpath /opt/tplink/EAPController/data/db
# run the mongoclient web client
docker run -itd --init --network mongonet --name mongoclient -p 3000:3000 mongoclient/mongoclient
For example: mongo.movIn this example, there is a user Once you're done, stop & cleanup the containers and restart the controller: docker stop -t 60 mongodb mongoclient
docker rm mongodb mongoclient
docker network rm mongonet
docker start omada-controller |
Beta Was this translation helpful? Give feedback.
-
|
Works! Had a duplicate entry for my admin user in tenant that I deleted. |
Beta Was this translation helpful? Give feedback.
-
|
I tried and it semi worked for me. I found the duplicate document containing my admin user with the wrong tenant_id and deleted it. I cleaned up and started the (still 5.9.31) container. It seemed to be running OK and I could connect trough the webinterface as well as the android app. So I upgraded to 5.12 and succesfully logged in using the web interface. However: the android app couldn't connect via my cloud connection or the local connection. Internet access also became intermittent and the Wife Acceptance Factor was quickly going down so I had to revert to the snapshot I took before... |
Beta Was this translation helpful? Give feedback.
-
|
Android app worked fine for me. I upgraded to 5.12 first and then did the mongodb workaround. Not sure if that makes a difference or not |
Beta Was this translation helpful? Give feedback.
-
|
I have a very similar issue. I had duplicated tenants, but I also have a failure in fetching the default omadac |
Beta Was this translation helpful? Give feedback.
-
|
This resolved my login credential issue. Now it says all my devices are 'MANAGED BY OTHERS' and I cannot adopt them with the same (and only) credentials. Anyone experience this and have any advice? |
Beta Was this translation helpful? Give feedback.
-
|
I have this issue, but my installation is on Raspberry. I tried those commands but it didn't work: When I add -d arm64v8/mongo to the "docker run" cmd, it runs ok, but when I go to the web page http://homeassistant.local:3000/ I don't get any answer. Any help on debugging that would be much appreciated :) |
Beta Was this translation helpful? Give feedback.
-
|
From what I can see there, it looks like a clean startup of a container with no previous persistent data. |
Beta Was this translation helpful? Give feedback.
-
|
Still not fixed in the current version? If so, that is ridiculous... I'm wondering whether Omada was the right choice, if they are not able to fix such significant issue. |
Beta Was this translation helpful? Give feedback.
-
|
My issue had more to do with how I ran this container, not the container itself. Furthermore, it's not the Omada issue, but how the controller is run. |
Beta Was this translation helpful? Give feedback.
-
|
I have been holding back from upgrading because of this. I switched from Unifi Ubiquity to TP-Link Omada because Unifi was releasing new fimrwares at such a rate that they often introduced more bugs than they fixed. I appreciated TP-Link's more conservative release cycles, but they seem to have gotten something wrong here... Anyway, I'm assuming that this issue will be closed once they fixed it, right? |
Beta Was this translation helpful? Give feedback.
-
|
I just asked again on the TP-Link forums post to see if it has been fixed or not but yes, I will update this issue if I ever get a reply. Their prioritization of issues and speed at which they sometimes fix them is annoying at times as significant bugs take months and months to get fixed. |
Beta Was this translation helpful? Give feedback.
-
|
I've just posted this reply in the TP-Links forums post:
|
Beta Was this translation helpful? Give feedback.
-
|
Me too! 🎉😁 |
Beta Was this translation helpful? Give feedback.
-
|
This might be a stupid question, but how do I update to the newest version without root access? Is there a better way than backup and reinstall completly? |
Beta Was this translation helpful? Give feedback.
-
|
Not totally sure what you mean by root access - if you're talking about in the container, you don't upgrade inside the container. You stop, remove, and re-create the container, attaching the persistent data to the new container. See this note in the readme about making sure to allow the container to gracefully shutdown to prevent database corruption. Make sure you have backups (either automatic through the software and/or take manual backups of the persistent data directories when the controller is stopped). |
Beta Was this translation helpful? Give feedback.
-
|
Allright. I thought there might be a better/faster way. I'll do this now, thank you. |
Beta Was this translation helpful? Give feedback.
-
|
Sure thing. The updating the container is the better, faster way 😉 |
Beta Was this translation helpful? Give feedback.
OK, so if someone who is comfortable doing some semi-advanced steps could try this out, let me know:
omada-controller/mydata/omada-controller/datauid:gidof your controller container is 508:508# stop the controller gracefully docke…