10 months ago
After auto upgrade to Railway Metal, already deployed tables in Postgres Database which were in production is not available. How can the tables be retrieved back. My Application is running in production
22 Replies
10 months ago
Can you please take this on priority. All the tables created on production are lost!!! Please help
10 months ago
Hello,
About 2 months ago you deleted the Volume for Postgres, without a Volume, data cannot persist.
Unfortunately, I cannot accommodate this request to restore data, going forward, please use a volume to store data.
For Postgres, you would want to mount the volume to /var/lib/postgresql/data
Best,
Brody
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
I was having the data in the database before the migration and all the users were using app in production. How can you say that data cannot persist ?
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Post the auto migration, all the tables are seen dropped from the database. Please try to check the logs or backups of the database.
All the exceptions are occurred in the Postgres service after the auto migration happened.
10 months ago
You deleted the volume 2 months ago.
Thus, for 2 months data was being stored in temporary storage that does not persist long term.
Please see our documentation on Volumes -
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
I never deleted the volume if I understood correctly. Why the data was lost after the migration ?
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Can you please help to restore the data in any way possible. There should be some backups right. Production app is heavily impacted due to this.
10 months ago
Was there any automated snapshot/backup available prior to the upgrade event ?
10 months ago
Hi Team,
I understand that the data was stored in ephemeral storage due to no volume being mounted, and that persistence is not guaranteed in that case.
However, this database served production users, and the data was intact and live until the Railway Metal auto-migration. We did not redeploy or restart the container ourselves during that time.
Given the criticality, I am kindly requesting:
1. Can you check if any internal backups, container snapshots, or disk volumes are still available before the upgrade?
2. Was there a backup taken during or prior to the Metal migration process?
3. Is there any way to access container data pre-migration, even temporarily?
I’m aware of the ephemeral nature of storage without volumes, but since the data was available just before the migration, I’m hoping something can still be retrieved.
Please treat this as a high-priority request — it has impacted our production users significantly.
10 months ago
Hi team, can you please provide an update.
10 months ago
Hi,
I'm really sorry you suffered a data loss of this level. We cannot restore this because it's not there anymore.
You removed your volume two months ago which caused your database to persist to ephemeral (temporary) file storage, which has no persistence or backup guarantees. Given that you were the one who initiated the deletion, we cannot help you with this further.
Again, I'm sorry you lost data of such importance. The only thing I can extend to you as a goodwill gesture is $250 in Railway credits. Please let me know if you would kindly accept that.
Ray
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
I was not aware that the volume was deleted somehow. The data loss happened after the migration run, which was not initiated by me. Can you please check if there is any backup or container snapshots were taken before the migration. I don’t know what I can do now, as all the data is lost. 
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Hello,
Here is the activity entry showing that you had deleted the volume 2 months ago -
Given that you were the one who initiated the deletion, we cannot help you with this further.
The offer for credits is still available, Please let me know if you would accept them.
Best,
Brody
Attachments
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
So, what about the migration run which happened automatically without any notification ?
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Hello,
We sent out a notice email on May 3rd, 2025 to let our users know that there will be ongoing migrations starting in May.
From our email dashboard, it shows that the email was opened May 3rd.
Please see our documentation on the Metal migration for more information -
https://docs.railway.com/railway-metal
Best,
Brody
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
Ah ok, I accept the offer. Pls credit it.
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Hello,
The credits have been added to your workspace.
Again, I'm sorry you lost data of such importance.
Best,
Brody
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
Today I upgraded to PRO and downgraded to HOBBY. Can I get the money back which I spent ?
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Hello,
I've refunded the $10.27 your card was charged, you will see that returned to your original payment method in 5 - 10 days.
Best,
Brody
Status changed to Awaiting User Response Railway • 10 months ago
10 months ago
Thank you, and do you have any suggestions how to go ahead with the app, as all the data is lost. Is there any way the spring boot api responses are cached in your server ?
Status changed to Awaiting Railway Response Railway • 10 months ago
10 months ago
Is it possible to get at least the response data of below api call made
requestId:"ZhnLrB1zQLqPtO0hvVzGSw"
timestamp:"2025-05-17T02:18:43.006853715Z"
method:"GET"
path:"/api/predictions/leaderboard"
host:"ipl-prediction-production.up.railway.app"
httpStatus:200
upstreamProto:"HTTP/1.1"
downstreamProto:"HTTP/2.0"
responseDetails:""
totalDuration:355
upstreamAddress:"http://[fd12:f13d:8f5f:0:2000:1b:94e6:ded4]:8080"
clientUa:"Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Mobile Safari/537.36"
upstreamRqDuration:355
txBytes:4243
rxBytes:763
srcIp:"117.205.208.161"
edgeRegion:"asia-southeast1"
10 months ago
Hello,
We never read the request or response body, thus we do not have the data from HTTP requests.
Best,
Brody
Status changed to Awaiting User Response Railway • 10 months ago
7 months ago
This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!
Status changed to Solved Railway • 7 months ago