queues backlog and retry/lag times available now in graphql!! 👀 thanks!

J

John SpurlockOP•1/9/23, 3:04 PM

Interesting note:

# Beta. Queue backlog data with adaptive sampling. Queues that are not being written to, or read from, will not return data, even if they have a backlog.

# Beta. Queue backlog data with adaptive sampling. Queues that are not being written to, or read from, will not return data, even if they have a backlog.

# Beta. Queue backlog data with adaptive sampling. Queues that are not being written to, or read from, will not return data, even if they have a backlog.

# Beta. Queue backlog data with adaptive sampling. Queues that are not being written to, or read from, will not return data, even if they have a backlog.

JJohn Spurlock queues backlog and retry/lag times available now in graphql!! 👀 thanks!

E

Emilio•1/9/23, 4:55 PM

Community Champions always on the ball, finding new features before we even announce them

JJohn Spurlock Interesting note: `# Beta. Queue backlog data with adaptive sampling. Queues tha...

E

Emilio•1/9/23, 4:58 PM

When your Queue is inactive we don't have any way (currently) to look at your backlog and log information about it

EEmilio Community Champions always on the ball, finding new features before we even anno...

J

John SpurlockOP•1/9/23, 5:10 PM

As much as I enjoy a good hate on graphql now and then, its standard way of schema introspection is pretty great

J

John SpurlockOP•1/9/23, 5:13 PM

Do

queue

queue

handlers have exactly the same limits (cpu time etc) as a

fetch

fetch

handler would have in the same worker? More time for unbound vs bunded - or do they behave differently like DO

JJohn Spurlock Do `queue` handlers have exactly the same limits (cpu time etc) as a `fetch` han...

M

Matt Silverlock•1/9/23, 5:26 PM

Same limits

Rrelero90 Noob question: Can I have a consumer also act as a producer? Trying to have a co...

M

Matt Silverlock•1/9/23, 8:53 PM

Have you configured the bindings for the script that also acts as your consumer? Are you calling

env.YOUR_BINDING_NAME.send()

env.YOUR_BINDING_NAME.send()

env.YOUR_BINDING_NAME.send()

env.YOUR_BINDING_NAME.send() (and passing in

env

env

as an arg?). Posting your code would help to confirm

Rrelero90 Here you go

M

Matt Silverlock•1/9/23, 9:38 PM

Code looks good. Let me investigate / repro.

Rrelero90 Here you go

M

Matt Silverlock•1/9/23, 10:14 PM

Can you

Object.keys(env)

Object.keys(env)

Object.keys(env)

Object.keys(env) and paste the output? DM me if you prefer.

Ssamcfinan Hi all -- what's the difference between Queues and Pub/Sub? I might be confusing...

M

Matt Silverlock•1/11/23, 2:18 PM

Queues is ideal for reliable messaging — typically between services — think of it as a job queue, a buffer, or a lightweight RPC API. Great for many to one or one to one use-cases. Native Workers support is built in.

Pub/Sub is focused much more on the broadcast use-case — common in IoT or with real-time apps — and especially for scenarios where you need MQTT as a protocol.

Ssamcfinan Hi all -- what's the difference between Queues and Pub/Sub? I might be confusing...

M

Matt Silverlock•1/11/23, 2:18 PM

(What problem are you looking to solve? That will help guide. More docs coming on this as both products mature)

M

Matt Silverlock•1/11/23, 2:19 PM

GCP Pub/Sub is somewhat in-between both; Queues is “closer” to AWS SQS than GCP Pub/Sub.

Ssamcfinan Thanks for the explainer! I think Queues is what I'm looking for based on your d...

M

Matt Silverlock•1/11/23, 2:20 PM

Definitely Queues. Pub/Sub won’t give you any controls to batch. Easy to do that in Queues.

Ssamcfinan Thanks for the explainer! I think Queues is what I'm looking for based on your d...

M

Matt Silverlock•1/11/23, 2:21 PM

https://developers.cloudflare.com/queues/examples/send-errors-to-r2/

Cloudflare Queues - Queues & R2 · Cloudflare Queues

The following Worker will catch JavaScript errors and send them to a Queue. The same Worker will receive those errors in batches and store them to a log file in …

M

Matt Silverlock•1/11/23, 2:21 PM

and: https://developers.cloudflare.com/queues/learning/batching-retries/

Batching and Retries · Cloudflare Queues

When configuring a consumer Worker for a queue, you can also define how messages are batched as they are delivered.

M

Matt Silverlock•1/11/23, 2:21 PM

Let Queues do as much of the batching for you as it can

Rrelero90 Perhaps a silly question but...What is the best way to go about combining multip...

E

Emilio•1/11/23, 5:03 PM

Valid question. I think a code example will best demonstrate how to approach this.

export default {
  ...

  async queue(batch: MessageBatch, env: Env, ctx: ExecutionContext) {
    if (batch.queue === 'my-first-queue') {
        doSomething()
    } else if (batch.queue === 'my-other-queue' {
        doSomethingElse()
    }
    ...
  },
};

export default {
  ...

  async queue(batch: MessageBatch, env: Env, ctx: ExecutionContext) {
    if (batch.queue === 'my-first-queue') {
        doSomething()
    } else if (batch.queue === 'my-other-queue' {
        doSomethingElse()
    }
    ...
  },
};

export default {
  ...

  async queue(batch: MessageBatch, env: Env, ctx: ExecutionContext) {
    if (batch.queue === 'my-first-queue') {
        doSomething()
    } else if (batch.queue === 'my-other-queue' {
        doSomethingElse()
    }
    ...
  },
};

export default {
  ...

  async queue(batch: MessageBatch, env: Env, ctx: ExecutionContext) {
    if (batch.queue === 'my-first-queue') {
        doSomething()
    } else if (batch.queue === 'my-other-queue' {
        doSomethingElse()
    }
    ...
  },
};

You can key off the name of the Queue in the

batch

batch

batch

batch object to know which Queue you're consuming from in that particular execution.

Rrelero90 I'm a bit stuck in a loop currently. Trying to delete a worker that was the cons...

E

Emilio•1/11/23, 7:51 PM

You need to remove the binding. You can do this in the Dashboard (

Workers > Overview > (your worker) > Triggers > Queues (bottom of page)

Workers > Overview > (your worker) > Triggers > Queues (bottom of page)

Workers > Overview > (your worker) > Triggers > Queues (bottom of page)

Workers > Overview > (your worker) > Triggers > Queues (bottom of page)) or by removing it from the

wrangler.toml

wrangler.toml

configuration and publishing the changes

Rrelero90 You are the literal most helpful today @oilylime !!!

E

Emilio•1/11/23, 7:54 PM

My pleasure, thanks for using our product!

M

Matt Silverlock•1/12/23, 2:28 PM

If we moved to support the ability to explicitly acknowledge (“ack”) messages within a batch, should the default behavior change to require you call

.ack()

.ack()

.ack()

.ack() on every message (or the entire batch) when handling it?

.retry()

.retry()

.retry()

.retry() would exist for the cases where you want to explicitly, negatively acknowledge.

If yes (always require an explicit ack),

this post; if no, reply with why! Keep in mind I am familiar with other messaging systems; there’s not an agreed-upon approach here!

MMatt Silverlock If we moved to support the ability to explicitly acknowledge (“ack”) messages wi...

M

Matt Silverlock•1/12/23, 2:29 PM

(This would be a breaking change for existing beta users unless we grandfathered today’s behavior)

Rrelero90 Question about how consumers are able to interact with a queue: Does a consumer ...

M

Matt Silverlock•1/12/23, 8:11 PM

Today: consumers are effectively single threaded. No concurrency. If your rate of production >>> your consumer completion rate, you’ll generate a backlog. Best practice right now is to have a single Worker consumer per queue vs multiple queues aimed at the same consumer script.

Near future: automatic consumer concurrency that scales out based on consumption rate

Aamitarelli Hi all - how do I access a queue directly from the outside web, without invoking...

M

Matt Silverlock•1/12/23, 8:12 PM

You’ll need to write a Worker - accept a HTTP request or even a WebSocket —> call

YOUR_QUEUE.send()

YOUR_QUEUE.send()

YOUR_QUEUE.send()

YOUR_QUEUE.send(). In the longer term we’re exploring other on-ramps to support this natively

FFlorian I would like the default behavior be to always manually acknowledge a message or...

M

Matt Silverlock•1/13/23, 12:20 PM

Can you expand? So you want to be required to ack() every message, or the batch, but any left unacked on an error response also don’t get retried?

Aaltryne hey, on this topic, I have a very long GPU process (~10 minutes) , and I've basi...

M

Matt Silverlock•1/13/23, 12:23 PM

The consumer will eventually hit running time limits and/or (very likely) connection limits if you are trying to keep a HTTP connection open for several minutes. I’d think about how you can instead have the single threaded machine - your GPU - poll for work when it’s ready vs blocking on it for minutes at a time.

U

Unsmart•1/13/23, 12:24 PM

I personally like the way AWS does it. By default all messages are assumed to be successfully processed if the function does not error or return a list of message ids that should be retired. And if the function throws an error all messages are retried.

U

Unsmart•1/13/23, 12:25 PM

Just nice to not have to think about needing to do something to messages you don't want to retry

UUnsmart I personally like the way AWS does it. By default all messages are assumed to be...

M

Matt Silverlock•1/13/23, 12:49 PM

Right, that’s the default behavior (unless you enable client acknowledgment mode) and prevents big retry loops if you forget to call ack(). OTOH you can silently lose messages if you don’t expect it.

U

Unsmart•1/13/23, 12:52 PM

Ah so both modes will exist you're just asking about which should be the default?
I'd still say what it is currently over manual requiring acks as the default.

UUnsmart Ah so both modes will exist you're just asking about which should be the default...

M

Matt Silverlock•1/13/23, 12:55 PM

I’m not promising anything just yet

Right now we think that the most intuitive approach is:

1. Messages are implicitly ack’d (as they are today) if the handler returns successfully. Users with simple use cases don’t need to do anything special.
2. You can call

.ack()

.ack()

.ack()

.ack() to make partial progress through a batch so that message is not retried if the handler errors
3. (Exists today) you can call

retry()

retry()

retry()

retry() or

retryAll()

retryAll()

retryAll()

retryAll() on a message or a batch to force a retry, even if the handler returns successfully

D

dave•1/13/23, 3:13 PM

Any recommendations on how to rate limit how often my queue can be processed? i.e. I’m working with a third party API that only allows X amount of operations per minute, so the queue will need to slow down firing off Workers during peak traffic.

U

Unsmart•1/13/23, 3:18 PM

Would probably be better to use DOs and alarms directly if you want that much control.

queues backlog and retry/lag times available now in graphql!! 👀 thanks!

Similar Threads

Similar Threads

Similar Threads