82 Replies
Project ID:
N/AWhat's this?
it's a message from percy asking you to provide your project id, you can get the project id from your url when you visit a project, it'll be something like
railway.app/project/PROJECT_ID
also #🛂｜readme #5fragly it's fine lol
I thought you read help threads
a project id is a unique ID that every railway project gets, and provides a way for the team to look into your project if you are having issues with the platform, you are just asking for help, so you can just send n/a
N/A
Do I just create serpbear from your template?
same username and password as used on pikapods
?
lets slow down a bit
do you have the sqlite database from your serpbear deploy on pikapods
Not sure
Where do I get it?
can you grab it via sftp like you did for kuma
ok
let me see
grab all those files
Done
okay now please give me a few minutes to get my dashboard cleaned up, go so many random project i dont know whats going on
ok
let me know
might be a good time for you to go through your projects and name the projects accordingly and delete anything that isnt in use
I did
All good
im not done lol sorry
ok
okay go ahead and deploy serpbear from the dashboard
with same username and password I have on pikapods?
ok
then deploy filebrowser into that project
filebrowser?
remove both services deployments
delete the volume from filebrowser
on filebrowser set USE_VOLUME_ROOT to 1
we remembered this time!
Yeah haha
done
disconnect the volume from serpbear and mount it to filebrowser at
/data
redeploy filebrowser
Done
did you delete any files yet
No
delete settings.json, then drop all the files you got from sftp on pikapods into that folder
into root?
yep!
done
screenshot?
looks good
remove the filebrowser deployment.
disconnect its volume
mount the volume to serpbear at
/app/data
redeploy serpbear
wait for a minute
login to serpbear with the same credentials you used on pikapods serpbear
let me try to reproduce this
ok
okay im so sorry, the username / password is not saved in the sqlite database, you need to use the username / password you set in the variables
ok
I'm in thanks
all looks good?
Yes
awsome
Bought you 2 more trains
thank you so much!
I am ashamed this is needed. I can't say I will fix this but is this truly the only way to load files onto a volume?
yep lol
well not strictly with filebrowser, any filebrowser will do, but the methodology is going to be the same
yeah, that's what I thought, if we allowed you to ssh into the service I guess you could install sftp in the docker image and use that with any sftp client.
Or something built into the UI
These are my ideas and not a roadmap for railway*
that's true, but probably the best way would be to natively give the user sftp access to the volume
True rather than relying on every template maintainer to keep an up-to-date version of sftp
still, that would be like wrapping any deployment in a meta deployment to accomplish that. Then you could flip a switch and expose ssh and sftp creds in the vars of the service.
Interesting idea. I will post it to the public roadmap and we can see if people want it.
personally I'm down for native sftp volume access, but not ssh access
Switches for both then? but sftp without ssh is just purposely restricted ssh. Personally, I think if you use public and private keys. There is nothing wrong with public-facing SSH.
personally I don't like the idea of ssh at all, switched or not
containers are empherial after all, ssh'ing into the container goes against that for the most part
The idea would be to create a tool to access your volume.
But I get your meaning, I think we would make this a very much turn on this switch when all else fails.
if ssh is ever offered at all, there are reasons to avoid it.
my 50 cents but isn't it easier to just make a microservice to access volumes data over https? just basic CRUD, something like cloudflare does for accessing r2 from workers
or add an S3 API to 
or SFTP or multiple instances with the same volume
then you can attach whatever you want to it
no actually adding an S3 api would be the best way to solve it I think
no need to reinvent the wheel and it's pretty easy to use
I think something more generic like sftp would be best, also not reinventing the wheel
wait can rsync work over stfp?
not sure but rsync isn't really the intended usecase
ok that's fair
but keep in mind this is only volumes v1, we have yet to see volumes v2
yeah if v2 will allow for connecting to multiple services at the same time - problem solved
can have that s3 as much as I want
that definitely makes it easier and less of a dance, but without native direct volume access it would still feel like a bit of a hack
Is there an automatic backup on railway or do I have to do manual backups?
Is ghost and other templates update automatically to their latest versions or do I have to do it manually or how does it work?
Railway doesn't have managed databases, so you'll have to perform manual backups
if they deploy from
image:latest then they'll always be updated to the latest image
just check where they deploy fromslight correction.
if they deploy from
image:latest then they'll always be updated to the latest image upon re-deploying the service
just check where they deploy from