Unfortunately Pages Functions doesn’t support node_compat bundling, nor the hyperdrive binding, so I

Unfortunately Pages Functions doesn’t support node_compat bundling, nor the hyperdrive binding, so I’m not aware of a simple way to get hyperdrive working with Pages, sorry

James•10/18/23, 4:03 AM

[ERROR] A request to the Cloudflare API (/accounts/cloudflare_account_id/hyperdrive/configs) failed.

not_entitled [code: 10000]

i got this error while using wrangler hyperdrive create name --connection-string="connection_string"

JJames [ERROR] A request to the Cloudflare API (/accounts/cloudflare_account_id/hyperdr...

James•10/18/23, 4:26 AM

anyone

JJames [ERROR] A request to the Cloudflare API (/accounts/cloudflare_account_id/hyperdr...

Sam•10/18/23, 9:03 AM

Are you on the paid workers plan?

SSam Are you on the paid workers plan?

James•10/18/23, 9:17 AM

Yes

JJames [ERROR] A request to the Cloudflare API (/accounts/cloudflare_account_id/hyperdr...

Matt Silverlock•10/18/23, 12:30 PM

Can you share your account ID please? (it's safe to share publicly)

MMatt Silverlock Can you share your account ID please? (it's safe to share publicly)

James•10/18/23, 12:32 PM

1d1c3f97856c67d63a1d69697bcb3b0d

JJames 1d1c3f97856c67d63a1d69697bcb3b0d

Matt Silverlock•10/18/23, 12:34 PM

You do not have a paid Workers plan.

Matt Silverlock•10/18/23, 12:34 PM

Hyperdrive (currently) requires a paid Workers plan per https://developers.cloudflare.com/hyperdrive/get-started/#prerequisites / https://developers.cloudflare.com/workers/platform/pricing/#workers

MMatt Silverlock You do not have a paid Workers plan.

James•10/18/23, 12:35 PM

this is the test account where we test the things before the production

James•10/18/23, 12:35 PM

so i think we need to login with that account through wrangler

JJames so i think we need to login with that account through wrangler

Matt Silverlock•10/18/23, 12:36 PM

Yep. You can use

CLOUDFLARE_ACCOUNT_ID

CLOUDFLARE_ACCOUNT_ID

as an env var too: https://developers.cloudflare.com/workers/wrangler/configuration/#inheritable-keys

Configuration - Wrangler · Cloudflare Workers docs

Wrangler optionally uses a wrangler.toml configuration file to customize the development and deploying setup for a Worker.

MMatt Silverlock Yep. You can use `CLOUDFLARE_ACCOUNT_ID` as an env var too: https://developers.c...

James•10/18/23, 12:37 PM

thx it's works

Huxwell | Reown•10/20/23, 11:18 AM

Hey team, I'm integrating Hyperdrive on an existing project and it works great.
However when I run the tests using unstable_devunstable_dev my tests are failing with the following error.
Any ideas what I could be doing wrong ? Here's how I init the test worker.

await unstable_dev(entry, {
      experimental: { disableExperimentalWarning: true },
      config: 'wrangler.toml',
    })

await unstable_dev(entry, {
      experimental: { disableExperimentalWarning: true },
      config: 'wrangler.toml',
    })

Which results in the following error.

> Unexpected token F in JSON at position 0
>  ❯ parseJSONFromBytes node_modules/wrangler/wrangler-dist/cli.js:6582:19
>   6580|     __name(utf8DecodeBytes, "utf8DecodeBytes");
>     6581|     function parseJSONFromBytes(bytes) {
>     6582|       return JSON.parse(utf8DecodeBytes(bytes));
>        |                   ^
>     6583|     }
>     6584|     __name(parseJSONFromBytes, "parseJSONFromBytes");

> Unexpected token F in JSON at position 0
>  ❯ parseJSONFromBytes node_modules/wrangler/wrangler-dist/cli.js:6582:19
>   6580|     __name(utf8DecodeBytes, "utf8DecodeBytes");
>     6581|     function parseJSONFromBytes(bytes) {
>     6582|       return JSON.parse(utf8DecodeBytes(bytes));
>        |                   ^
>     6583|     }
>     6584|     __name(parseJSONFromBytes, "parseJSONFromBytes");

HHuxwell | Reown Hey team, I'm integrating Hyperdrive on an existing project and it works great. ...

Hello, I’m Allie!•10/20/23, 11:21 AM

AFAIK unstable_devunstable_dev is local-only, so it might be failing because Hyperdrive doesn't work locally?

meai1•10/22/23, 2:29 AM

Will hyperdrive support Microsoft Sql Server at some point?

meai1•10/22/23, 2:33 AM

also I don't really understand how or why hyperdrive would make uncached queries faster.. it doesn't make sense. Especially not " 6 - 8x faster on uncached queries"

kian•10/22/23, 2:36 AM

The blog where I assume you got those numbers from explains it all - https://blog.cloudflare.com/hyperdrive-making-regional-databases-feel-distributed/

meai1•10/22/23, 2:37 AM

so merely connection pooling is doing this? But I think most poeple of course use connection pooling without cloudflare too

kian•10/22/23, 2:42 AM

Hyperdrive is specifically for Workers, which are stateless and generally short-lived.

kian•10/22/23, 2:43 AM

You can’t hold a connection open in a Worker, it is scoped to that request. Hyperdrive gives you regional connection pools.

kian•10/22/23, 2:43 AM

https://blog.cloudflare.com/workers-tcp-socket-api-connect-databases/ talks about it briefly, as the TCP connect API is what Workers (and database drivers for Workers) use to talk to databases.

kian•10/22/23, 2:45 AM

There’s also the Cloudflare backbone, which you can use yourself generally with Argo Smart Routing, which is self-described as Waze for the internet.

It’s constantly probing around and finding the best network path from A to B, and usually results in a pretty large latency decrease for cross-region requests.

kian•10/22/23, 2:46 AM

In the same region, sure, ASR is more like 5-7% - but we saw improvements of 55% from US and APAC to an origin in the UK,

Add in connection pooling, and query caching, and you have Hyperdrive.

Sam•10/23/23, 5:47 PM

Probably asked before, but couldn’t find it in the chat history:
Is there any ETA of this becoming General Available? So out of beta? Thanks.

SSam Probably asked before, but couldn’t find it in the chat history: Is there any E...

Hello, I’m Allie!•10/23/23, 5:47 PM

SSam Probably asked before, but couldn’t find it in the chat history: Is there any E...

Matt Silverlock•10/23/23, 7:21 PM

It’s about ~4 weeks into beta, so won’t be GA until next year. Hold tight

MMatt Silverlock It’s about ~4 weeks into beta, so won’t be GA until next year. Hold tight 😉

Sam•10/23/23, 7:22 PM

Ahh I thought it was in alpha and beta for a year now, thanks. Can't wait for GA!

SSam Ahh I thought it was in alpha and beta for a year now, thanks. Can't wait for GA...

Matt Silverlock•10/23/23, 7:25 PM

Do you perhaps mean D1? Hyperdrive was only announced ~4 weeks ago when we launched it - never been mentioned before that.

MMatt Silverlock Do you perhaps mean D1? Hyperdrive was only announced ~4 weeks ago when we launc...

Sam•10/23/23, 7:25 PM

Yup misread it, forgot to select the "in:hyperdrive-beta" filter while searching for any previous messages

Sam•10/24/23, 2:57 PM

I'm aware of it being a beta, and no I'm not currently using it in production. But I would like more information about these following to limits. What do they exactly mean?

For example, if my worker gets called 1,000 times per second (that would be 1,000 Hyperdrive connections). Would this be possible with these current limits?

Unsmart•10/24/23, 3:27 PM

The number of connections depends on the number of regions serving your traffic... if all 1000 requests hit a single region it would be at most 10-20 connections used to the origin.

Sam•10/24/23, 3:35 PM

Ahh ok

anirudh•10/26/23, 1:46 PM

Is there a way to limit the number of connections Hyperdrive uses to connect to the origin DB?

anirudh•10/26/23, 2:02 PM

I know that it takes a max of 10-20 per region and overall it will be a max of ~100, but my DB is on DigitalOcean and they allow only 40 connections

Aanirudh Is there a way to limit the number of connections Hyperdrive uses to connect to ...

Matt Silverlock•10/26/23, 4:48 PM

Not yet.

Colin•10/26/23, 5:03 PM

Hey, curious if the changelog for Hyperdrive (or more generally, Cloudflare products) includes bug fixes/stability improvements, or just new features. We're currently holding off on using Hyperdrive in dev until a couple routing issues I discussed w/ alex robinson are able to be improved, but want to know roughly when we can start re-evaluating Hyperdrive in test environments

CColin Hey, curious if the changelog for Hyperdrive (or more generally, Cloudflare prod...

Matt Silverlock•10/26/23, 5:19 PM

We’ll let you know. We include some bug fixes in changelogs - it really depends on the impact/breadth.

Colin•10/26/23, 5:33 PM

ah fantastic, thank you so much!

Colin•10/27/23, 8:25 PM

hi again y'all - we're seeing a large number of "could not get connection from the pool" out of our Postgres driver (pgpg)

Colin•10/27/23, 8:26 PM

This seems to happen in a very brief spike, and then go away

Colin•10/27/23, 8:26 PM

Seems to happen more during lower volume periods

Colin•10/27/23, 8:26 PM

We're not seeing any underlying issues on the DB (or from clients connecting directly), only via Hyperdrive conns

CColin hi again y'all - we're seeing a large number of "could not get connection from t...

Matt Silverlock•10/28/23, 7:20 AM

Can you share the account ID and the Hyperdrive config ID?

MMatt Silverlock Can you share the account ID and the Hyperdrive config ID?

Colin•10/28/23, 10:15 AM

Yep.

CF account: cfd355bcff7e6af6838743ae944184b0
Hyperdrive: lunarclientprod-com

CColin hi again y'all - we're seeing a large number of "could not get connection from t...

Matt Silverlock•10/28/23, 12:24 PM

Can you also share your JS code? You’re not using PG’s “Pool” class, right?

MMatt Silverlock Can you also share your JS code? You’re not using PG’s “Pool” class, right?

Colin•10/28/23, 1:43 PM

Ah, I am using PG’s Pool. Is there a reason that isn’t preferable w/ Hyperdrive?

CColin Ah, I am using PG’s Pool. Is there a reason that isn’t preferable w/ Hyperdrive?

Matt Silverlock•10/28/23, 2:30 PM

It’s client side, which won’t work: the pooling has to exist outside of the function.

Matt Silverlock•10/28/23, 2:30 PM

(A big reason why Hyperdrive exists!)

MMatt Silverlock It’s client side, which won’t work: the pooling has to exist outside of the func...

Colin•10/28/23, 2:32 PM

So we were only using PG’s Pool to support using >1 conn (to Hyperdrive) as we make a large amount of parallel queries. The PG pool isn’t kept in scope past 1 worker request or anything, so I just viewed it as an easy way to achieve parallelism to Hyperdrive (which would then do the same to the DB)