My table that was broken seems like it still is so maybe its not fully rolled back yet or they redid it not sure. Not too worried as I wouldnt use anything production with a beta product and it happened on a table I dont even use and I dont know why I made
Fully rolled back. A single table being broken is something else. The issue prevented the DB from loading. There is no such thing as a “partial” load - we don’t load incremental pages. I would not associate with today’s issue.
I meant database not table in that message... I dont think it even has any tables but not really sure since I dont use that database at all and dont remember making it
It's not the first time folks have hit this same error message before. I'm not sure if the root cause is the same or not, but especially for issues like this that externally seem to be recurring ones, it would be really appreciated if a status page could be put up in future. It saves a lot of the reactive back and forth that happens here, and only improves the transparency we all know and love Cloudflare for
A package to dump a copy of your experimental D1 database through a worker. - GitHub - nora-soderlund/cloudflare-d1-backups: A package to dump a copy of your experimental D1 database through a worker.
(not a team member, just from my knowledge as an outsider) First: The goal is still to increase this during the beta. Second: With D1, you can have a lot of seperate databases, all with a limit of 2gb. If your application can support sharding the DB like this, there isn't really a limit to the amount of data you can store. To me it seems perfect for a lot of SaaS applications, where each customer, or even each project, would have their own database.
so if a client would then need to make a query, the query would have to first look at which database to look into and perhaps query the data across databases?
a certain issue manager that many folks dislike handles multi-tenancy that way, yeah - one db per customer.
you keep a highly available store mapping something (customerId, a subdomain, etc) to the database ID, then pass the databaseID to your client (we're working on making the d1-api support this nicely)
but if a customers needs way more than 2gb of data, then do you see this viable? I am trying to setup a flow where a js script on websites sends events data to a cloudflare worker, from which I am trying to understand how to best forward it to a database
in the dashboard, is there a plan to add rows read/written so I can have an idea on what my bill would be? I know I can see it for the past week by going to the per-db metrics, but I would like to be able to see the total for the month since that what i'll be billed for once billing gets enabled