A very interesting alt method. I may consider this, though I doubt whether it can handle frequently
A very interesting alt method. I may consider this, though I doubt whether it can handle frequently CRUD.
CIt can handle it a lot better then KV does. KV is eventually consistent and should only be one write per key per second
Chttps://developers.cloudflare.com/r2/reference/consistency/ R2 is read after write/strongly consistent.
Although it may be worth mentioning: Why did you say Durable Objects has some trade offs you don't like? R2 is backed by DOs for metadata and then stores the actual objects elsewhere
Although it may be worth mentioning: Why did you say Durable Objects has some trade offs you don't like? R2 is backed by DOs for metadata and then stores the actual objects elsewhere
BFor my next purpose, I want to use it for de-duplicate. Each of my ID is 128-byte, and the durable objects' size limit is 128KB, so only 1k records will exceed the limit, which will happen in less than one second. Even the durable objects' with SQLite backend can be filled with only 16k records. Since I need to store at least 7 days data, it's far from my need.
BOh, that's first time I know that some part of KV is backed by DO.
But, from the document, I don't find any related to the cache time around the metadata in KV. I guess the cache policy is the same as value, right?
But, from the document, I don't find any related to the cache time around the metadata in KV. I guess the cache policy is the same as value, right?
Cahh ok, yea the max key size for R2 is 1,024 bytes and there's no bucket max
Oh, that's first time I know that some part of KV is backed by DO.R2 is, not KV
But, from the document, I don't find any related to the cache time around the metadata in KV. I guess the cache policy is the same as value, right?For KV yea should be the same, but that document I linked was about R2, not KV
Nhi. beginner here. i'm having trouble with accessing KV from a worker. i did the binding but keep getting errors
Nkeep getting "Property 'get' does not exist"
Nin the worker code, should i refer to the name or to the value of the binding?
Nwould appreciate any assistance
CWhatever you set
bindingenvCloudflare Docs
KV bindings allow for communication between a Worker and a KV namespace.
CIf you're using service worker format
addEventListener('fetch',... (migrate to module workers is recommended) then it's a global varNif i have that, which should i use for .get and .put, the name or the value?
Cwishlist_kv, the nameNif i did that binding to the worker in the ui, do i need to declare or initialize that kv var before using it?
wishlist_kv.get(key) gives me Cannot find name 'wishlist_kv'.Nand
Property 'put' does not exist on type '"wishlist_kv"'Cno, show more code though. If you're using module workers/exporting fetch, it's on env and not in global, there's a nice tutorial here: https://developers.cloudflare.com/kv/concepts/kv-bindings/#access-kv-from-workers
Cloudflare Docs
KV bindings allow for communication between a Worker and a KV namespace.
Ncode starts with
addEventListener('fetch', event => {
event.respondWith(handleRequest(event.request));
});
Ndo i need to add env there?
Cno, you're using the old only supported for some products service worker format which should have it as a global var
Ni'll have to read up on migrating to module workers?
CKV should work fine with service workers but module workers are a smarter thing to use for performance/more features/etc reasons, and not that hard if you're just starting to switch: https://developers.cloudflare.com/workers/reference/migrate-to-module-workers/
Otherwise would need to show more of the full code of the service worker
Otherwise would need to show more of the full code of the service worker
Cloudflare Docs
Write your Worker code in ES modules syntax for an optimized experience.
Nthanks
DI have some follow up questions for the same, can I DM you?
T
TOf course! DMs always open
AHi, I'm having some trouble expiring keys in KV, I'm using Rust and it's an open source project so you can see I set it here https://github.com/aquarhead/nextgame/blob/master/src/lib.rs#L502 I just enabled logs and according to it I should have set an example object to expire at 19:20 UTC (10min ago)
But it's still there even now. Am I missing something apparent here? (This seems to work correctly locally in dev mode)
But it's still there even now. Am I missing something apparent here? (This seems to work correctly locally in dev mode)
GitHub
Who's playing?! Contribute to aquarhead/nextgame development by creating an account on GitHub.
AThe
Game should not exist after 19:20
Chey there.
I'm having two workers - one serving astro and another as a service auth worker
getting this response type when calling the service worker: ProxyStub { name: 'Object', poisoned: false }
I'm having two workers - one serving astro and another as a service auth worker
getting this response type when calling the service worker: ProxyStub { name: 'Object', poisoned: false }
Cstill, the response in the service worker (the user object) comes fine from the KV
Coh it looks i have to stringify when i'm getting the object
Swhere did you get the 16k record limit for DO sqlite? I've got like 4MB in there and def more than 16k records
DWe are suddenly experiencing a significant latency when updating a value in the KV. We are experiencing something like 30 seconds latency. Is there some known problem in the datacenter?
AI am having the same issue
DSo is this a known problem at Cloudflare? This causes big issues in our application
ADoesn't seem so, nothing on their status page
DYeah saw that too
Abut I think generally write latency can be up to a minute
but it usually isn't
but it usually isn't
DYeah it's usually more or less instant
DAre US or Europe your nearest location? Seems that cloudflare kv storage location is primarily in US and Europe. Maybe the one in Europe is down and rerouted to US, that would explain the issue
DBut would be good to get some input from Cloudflare
AI am in europe but I respond to discord interactions (for a discord bot) which might not come from europe
D
IWe are seeing the same latency spikes. Actually came here because there was no update on the status page.
JI'd like to flag this to the KV team - can you let us know whether you're using REST APIs or Binding APIs?
AI am using the binding API. I would also like to add that it's not only latency. For me in some cases the values are never written.
DBinding API here also. We noticed the problem this morning, about 6 hours ago.
JThanks both, I've passed on all the info to the KV engineering. Hopefully someone with more domain knowledge from CF will pick this up when they see this 
IJust to confirm that it doesn't look like isulated case we are seeing this on multiple deployments across regions.
TThanks everyone, we're looking into this
T@danielsundq @ivogt @Anthony can you share in this thread since when you've been encountering these issues?