Help Switching to JSON Redis Stack
Hello there, I'm trying to use a version of Redis on railway with the JSON extension. I came across Brody's answer here and started to follow along:
https://www.answeroverflow.com/m/1154090266093953065. Only issue is that the env variables didn't seem to carry over well. I had to change around some ports to get it to connect again but now it's not reading from the attached volume. All online help talks about changing the docker start command but the Build section is missing from the settings panel. How can I get it to connect to our existing volume?
Activate the JSON module on redis - Railway
When creating a redis db is there a way to activate the json module without creating a custom docker image but rather using the native railway way of creating a redis db?
I tried the MODULE LOAD command but it did not work.
Thanks!
38 Replies
Project ID:
684aa05b-448a-4194-93fb-fa5a9fc97cc4
684aa05b-448a-4194-93fb-fa5a9fc97cc4
first off, can you please confirm if the start command field is missing from the service settings? not that I think you would need it, but that's an issue worth tackling first
No the start command is visible
Just build is missing
So should I set the start command to:
docker run -v /local-data/:/data redis/redis-stack:latest
And change my mount path to /local-data
oh okay sorry I misread your message a tad, build command would not be applicable for docker images so it is not shown in the UI
you can not change the docker run command railway uses like that
Yeah sorry I was just confused because I deploy a docker image for another service on railway and the build section is present for that one
you have linked my previous thread that worked for them, I will try my method that I gave in that thread and let you know if it still works
Great. To be clear my main issue is it connecting to a persistent volume. Pref one that already exists
did you have data in it?
yes
I can see the data is still there by looking at the data usage metric on the attached volume. I just don't think the redis instance is connected to it
how much disk is in use?
~170mb
could just be the base amount of disk usage
redis is mostly an in memory store after all
Pretty sure it's our data. I'm quite positive the default railway redis deploy uses persistence
A brand new deploy is around 140mb
okay sounds good, I will first check the default redis database for data persistence, and then check if my original solution still works when back at my computer
Yes please do because we low key just fried this database
you could always swap the volume itself back over to a regular redis database deployment in the mean time?
How can I move volumes?
unmount and remount
if I get this working would you be interested in adding me to your project so I can make sure your redis database is setup correctly as it sounds like you may have made some unsavoury changes to it
I can confirm by switching the image back to bitnami/redis:latest and then changing the mount point on the volume back to /bitnami I was able to restore my db back to what it was before the attempted change to one with JSON
But this further proves that the main Redis image uses persistence
oh thats awesome, okay delete the redis json service then please as I'll provide instructions from scratch
Sounds good
hey @Marc its not going to be possible to swap the volume from bitnami's redis to redis stack, they use a different file structure.
so instead, would you mind adding me to your project so i can get a new redis stack deployed for you, then i will set you up with a set of commands that will copy the data over!
Sure how can I go about adding you?
my email is [email protected]
add me via the members tab of your project
Done
do you want to run the commands, or would you like me to?
You can give it a shot
will do
data has been replicated, do i have your permission to switch the linkedIn-api service over to redis stack?
I can probably take it from here. I don't want to keep taking up your time. You've been super helpful!
i would need to stop replication after its been switched over, thats why i asked if it was okay if i could switch it over
AHH
then yes
okay its redeploying now
can i take the Cache service offline? (this wouldnt remove any data)
yup
okay you should be good to go now, let me know if everything works as it should
Okay I'll test it rn
Seemingly we're all good! Thank you so much for your help!
no problem, happy to help!
feel free to remove me from the project!
Yup just did