Cloudflare Developers•3y ago

oh wait of course, there's do-taskmanager!

Jjohtso oh wait of course, there's do-taskmanager!

DaniFoldi•3/11/23, 6:18 PM

and do-transactional-outboxdo-transactional-outbox

johtsoOP•3/11/23, 6:20 PM

has anyone made a cloudflare worker / DO resources cheat sheet github repo or something?

johtsoOP•3/11/23, 6:20 PM

with all these things linked

johtsoOP•3/11/23, 6:21 PM

itty-durable, do-taskmanager, do-transactional-outbox..

Ffrankichiro Isn't miniflare part of wrangler? Do I need to install miniflare separately, or ...

Hello, I’m Allie!•3/12/23, 9:03 AM

There are three tiers.

wrangler

wrangler

is the CLI, miniflareminiflare is the system that actually mocks a Workers environment, and workerdworkerd, if you choose to use it, is the actual Workers runtime.

Advany•3/12/23, 9:58 AM

is there like a framework on top of do that can handle spinning of new do instances when limits are reached?

I want to save json objects inside of DO to be quickly available during requests. If I want to store like couple 1000 of json objects inside one DO (one do per client). Would I store everything in one object like this.configs = {} or is there a more efficient way of doing it?

Ffrankichiro So… getting DO to work with Pages just needs two CLI flags, but getting DO to wo...

Hello, I’m Allie!•3/12/23, 10:14 AM

No, I was just explaining what the different parts did. To use a DO with Workers, all you need is a

wrangler.toml

wrangler.toml

johtsoOP•3/12/23, 12:38 PM

"We do plan on improving the lifecycle management of objects soon, and auto-expiration is very likely to be part of that." saw this in a forum post from 2 years ago, did anything happen on this front?

Jjohtso "We do plan on improving the lifecycle management of objects soon, and auto-expi...

Hello, I’m Allie!•3/12/23, 12:39 PM

Can you link the post in question?

johtsoOP•3/12/23, 12:39 PM

https://community.cloudflare.com/t/durable-objects-larger-storage-limit-per-object-time-invalidated-storage/222156/13

Cloudflare Community

Durable Objects - larger storage limit per object + time-invalidate...

Not quite. The Durable Object is actually a special worker instance routed to by name/ID that has access to its own in-memory state and persistent state, the latter of which is exposed via a key-value interface. We do plan on improving the lifecycle management of objects soon, and auto-expiration is very likely to be part of that. I don’t have a...

johtsoOP•3/12/23, 12:40 PM

I guess this was before alarms

johtsoOP•3/12/23, 12:40 PM

And now alarms are basically the tool for all lifecycle related things?

Jjohtso I guess this was before alarms

Hello, I’m Allie!•3/12/23, 12:41 PM

This is probably the answer, from the thread:

You can build a TTL using alarms. Whenever your object is accessed, you would use setAlarm()setAlarm() to set the alarm time to Date.now() + timeoutDate.now() + timeout. There is only one alarm per object, so each call to setAlarm()setAlarm() replaces the previous alarm with the new alarm time. When that time is reached, the alarm handler will be invoked. In your alarmalarm handler, you would use storage.deleteAll()storage.deleteAll() to delete your object.

Since it’s possible to build TTL using alarms, we probably won’t be adding a separate TTL feature.

- Kenton

johtsoOP•3/12/23, 12:45 PM

Ah well spotted

johtsoOP•3/12/23, 12:46 PM

I guess better that energy is put towards stuff that can't be done in user land

SSubh Hey team, I removed one DO call and one Queue call, my api response time down fr...

Larry•3/12/23, 6:28 PM

I don't think you provided enough details to make a full diagnosis and even if you did, I have no experience with queues and I'm far from knowing all the ins and outs of DO performance. One wild ass guess though... do you use env.MY_DO_NAMESPACE.idFromName(name)env.MY_DO_NAMESPACE.idFromName(name)? Those require a global lookup on object creation and should be avoided.

johtsoOP•3/13/23, 1:12 AM

just found this too https://github.com/ticket-bridge/hyper-durable

johtsoOP•3/13/23, 1:14 AM

the extra layer of abstraction it proposes makes my head hurt though

Chaika•3/13/23, 4:04 PM

You can use location hints to "hint" that it should be within a specific region: https://developers.cloudflare.com/workers/runtime-apis/durable-objects/#providing-a-location-hint, as well as restricting objects to a specific jurisdiction (like for GDPR), but nothing direct to a specific colo. Not all colos support Durable Objects either (https://where.durableobjects.live/)

Jacob Wright•3/13/23, 5:15 PM

Does anyone know if a DO connects to another via websocket, will they both stay open until one of them closes the socket connection, or will they shut down after no external requests have been made to them?

Jacob Wright•3/13/23, 6:13 PM

Is there a possible race condition if 2 requests call idFromNameidFromName with the same never-before-used name? Or are there guards around that?

JJacob Wright Is there a possible race condition if 2 requests call `idFromName` with the same...

DaniFoldi•3/13/23, 6:18 PM

I've tested that and it was fine, the first one (1-2ms apart) got it, and if my memory serves me right, idFromName first checks with a global coordinator before "assigning" it to the closest colo, hence the increased perceived latency of the first request

DaniFoldi•3/13/23, 6:19 PM

I'm assuming all colos have a map of ids stored so subsequent accesses are faster, so it's just a first-time thing

DDaniFoldi I've tested that and it was fine, the first one (1-2ms apart) got it, and if my ...

Jacob Wright•3/13/23, 10:51 PM

So you think the global coordinator is able to ensure only 1 DO by name exists, even if 2 requests come in at the same time for a new name? Thank you for your feedback. I'll also be sure to test it, but that's good to know.
It's interesting idFromNameidFromName doesn't return a promise since it can be async.

JJacob Wright So you think the global coordinator is able to ensure only 1 DO by name exists, ...

Hello, I’m Allie!•3/13/23, 10:52 PM

That’s because the check doesn’t actually occur until the stub is fetched the first time

oh wait of course, there's do-taskmanager!

Similar Threads

Similar Threads

Similar Threads