-
Notifications
You must be signed in to change notification settings - Fork 25
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Datasets must have the same number of columns #860
Comments
In my case is very similar but on IPv6 training: {"reqId":"ZH7FOecwUjPZ1VFjfqIC","level":3,"time":"2024-03-06T22:20:01-03:00","remoteAddr":"","user":"--","app":"suspicious_login","method":"","url":"--","message":"Caught unknown error during IPv6 background training","userAgent":"--","version":"28.0.3.2","exception":{"Exception":"Rubix\ML\Exceptions\InvalidArgumentException","Message":"Datasets must have the same number of columns, 80 expected, but 16 given.","Code":0,"Trace":[{"file":"/var/www/nextcloud/apps/suspicious_login/lib/Service/DataLoader.php","line":131,"function":"merge","class":"Rubix\ML\Datasets\Labeled","type":"->"},{"file":"/var/www/nextcloud/apps/suspicious_login/lib/Service/TrainService.php","line":72,"function":"generateRandomShuffledData","class":"OCA\SuspiciousLogin\Service\DataLoader","type":"->"},{"file":"/var/www/nextcloud/apps/suspicious_login/lib/BackgroundJob/TrainJobIpV6.php","line":70,"function":"train","class":"OCA\SuspiciousLogin\Service\TrainService","type":"->"},{"file":"/var/www/nextcloud/lib/public/BackgroundJob/Job.php","line":81,"function":"run","class":"OCA\SuspiciousLogin\BackgroundJob\TrainJobIpV6","type":"->"},{"file":"/var/www/nextcloud/lib/public/BackgroundJob/TimedJob.php","line":102,"function":"start","class":"OCP\BackgroundJob\Job","type":"->"},{"file":"/var/www/nextcloud/lib/public/BackgroundJob/TimedJob.php","line":92,"function":"start","class":"OCP\BackgroundJob\TimedJob","type":"->"},{"file":"/var/www/nextcloud/cron.php","line":152,"function":"execute","class":"OCP\BackgroundJob\TimedJob","type":"->"}],"File":"/var/www/nextcloud/apps/suspicious_login/vendor/rubix/ml/src/Datasets/Labeled.php","Line":364,"message":"Caught unknown error during IPv6 background training","CustomMessage":"Caught unknown error during IPv6 background training"},"id":"65eb441600caa"} Every day at about 10 PM |
This comment was marked as duplicate.
This comment was marked as duplicate.
Same here. I run Nextcloud only through IPv6 (for other reasons).
|
Getting the same error for IPv4 Nextcloud Version: 28.0.4
Seems like this error only started recently since upgrading to the latest Nextcloud. |
This comment has been minimized.
This comment has been minimized.
I am also having this issue for IPv6, running Nextcloud 28.0.5.1 on Ubuntu 20.04.6 LTS server.
|
getting a similar one, selfhosing nextcloud aio behind a swag reverse proxy
|
This comment has been minimized.
This comment has been minimized.
Does anyone have an idea to what could be causing this issue ? Are the devs, wining ? This issue seems/looks to be reproducible. |
This comment has been minimized.
This comment has been minimized.
List the reproduction steps and I'll check it out. |
I think it might have something to do with the absence of enough training data
|
I have that same output, but when using IPv6:
Using IPv4:
|
Mhm, yeah, I have this feeling as well. After the error occoured on my instance, I've deleted all entries in the Right now, after a bit more then a week, I've got ~5500 recorded logins with 16 IPs/tupels. That's even more then I've had before cleaning up the database tables without any change in usage (just me and my father). 🤔 Maybe there is something broken with the data recording? Another information that may be interesting to the developers is the fact that I've disabled the plugin for quite some time in between. I had https://github.com/nextcloud/mail installed as well which lead the suspicious_login plugin broken due to an old RubixML version being loaded into memory. After identifying the issue and disabling/deleting the mail plugin, I've re-enabled the suspicious_login plugin. That means my datasets were not continuous. As one way to debug the error, I would personally start by inspecting the |
Without any changes this is in my tables:
Thats the issue, the tables look normal. The instance isn't really "broken" it just fails to log most of the login data. |
@ChristophWurst I had some time to setup a new VM and nextcloud instance. To reproduce the issue simply do a clean install (latest version). Install recommended apps. Enable suspicious login app.
Both commands return:
Meaning there is no model available. Can you perhaps provide a working model from your setup |
@theDepart3d "To reproduce the issue simply do a clean install." I am sorry, but you are not reproducing the error stated. The issue occurs when there is data present in the mentioned tables. Simply cleaning and/or starting fresh does not guarantee the issue will not occur, when data is being written to the mentioned tables |
@SjoerdV Interesting. The nextcloud instance that has the error, simply does not populate the tables. Thats where I have no clue why. The new VM is not publicly accessible, Ill get a domain later and leave it to populate the data. If it populates the data. The VM with the errors also show the same error "Not enough data, try again later". The instance is publicly available but it does not populate the tables. It uses a model but where does the default model get stored or does it create a model from the populated data ? |
@theDepart3d Yes, you have to wait (and actually use the instance), to have the tables populate. Do please review the original error as posted by OP. Happens during training on data collected and stored in mentioned tables |
This comment has been minimized.
This comment has been minimized.
The same problem.
And
I have 4 active users in my Nextcloud instance. I think it is safe to presume that this app is useless under certain amount of users since it will never generate enough login attempts to actually train model in some meaningful way. I think this app should be removed as default from Nextcloud. Also it would be very helpful if devs could put in main page information about minimum amount of data required to train model. |
I'm using the app with a three people instance without problems. If your users always log in from the same IP there won't be a lot of data indeed. But as soon as someone uses a phone to connect, there's some variance in IP addresses and plenty of data.
Sure: https://github.com/nextcloud/suspicious_login?tab=readme-ov-file#neural-net |
It's not a default app. It's shipped, but disabled by default. |
I use it with only 2 users and have no more problems so far after my "reset" described here. Over 100 IPs with ~40.000 logins were collected so far. But the plugin only collected enough informations to train the model and activate it for 8 days yet. I will report here if the problem re-appears. |
Could you please elaborate how did you solve the problem? I am using nextcloud in a docker container with only 2 users active. Seeing daily multiple errors in this regard. Thanks |
@kraizelburg you would need to access your Nextcloud DB and clear all the rows from each of the following tables: |
This comment has been minimized.
This comment has been minimized.
Still an issue, try to remove the app, purge the table with the problem then re-enable the app, I still got the issue for IPv4 and IPv6. I think I should drop the table but not sure how to force the app to recreate it after that |
It's been ~3 months since I cleared the associated tables in the DB, and I have not seen this issue since. I'm currently running the latest Nextcloud Docker image (Nextcloud Hub 9 30.0.1). My training data shows 409144 logins over 72 distinct IP's since I cleaned the tables. I see some people are still facing this issue, but for me, clearing the tables as described above seems to be an effective solution. |
I am getting the following error in Nextcloud. Could anybody help me?
The text was updated successfully, but these errors were encountered: