You can also just create new “namespaces” by giving each a prefix
You can also just create new “namespaces” by giving each a prefix
BYeah this is a thing, but that workaround wouldn't be good in combination with Workers for Platforms, otherwise the user workers will be able to see other's data
BI've never tried the limit increase before but I've seen some people here mention their requests went unanswered for weeks so it puts me off a bit (and it is strange hosting it on a Google form, doesn't feel official)
MMMfigure out how big your rows are, divide 10GB by that size, that's how many rows you can store
Mthe classic challenge is figuring out the right way to split the data, not always easy
Mnot without polling the db, not yet anyway
Msay you have a blog with 10kb of text (~1k words) per article, that's a million articles for 10GB
Mthis doc might be useful https://developers.cloudflare.com/d1/build-with-d1/query-json/
HCould you pass in an RPC target that makes reads only to their “namespace”?
Mwouldn't you want to do something like:
json_extract(titles, $.romanji) as romanji, json_extract(titles, $.english) as english, ... ?Mnot sure if sqlite supports the idea of nesting objects in a response
Mtitles -> $ or titles ->> $ might do it
Dis it safe to leak the database id? i.e. to publish
wrangler.toml
HYes. The entirety of
wrangler.toml
AThe ID only works for workers in your account so it should be fine
MMy where condition with an index on id is taking 2.0-2.2 seconds. If I get everything from the table it takes 2.3-2.8 seconds. Is it normal?
M
M
Many thoughts? thanks!
MMOops
Mokay, is it good? 500+ms...
Mit ranges from 500+ms to 1.2 seconds
