hi, hi 👋 Thank you for reporting these potential issues (and @HardlyWorkin' for raising it). I m

M

Matthijs•2/9/24, 8:27 AM

Hi, we are seeing the same for all of our D1 databases.

M

Matthijs•2/9/24, 8:27 AM

Located in Europe (AMS)

M

Matthijs•2/9/24, 8:30 AM

@ale
Here's what the API is returning:

{
  "result": null,
  "success": false,
  "errors": [
    {
      "code": 7501,
      "message": "Internal error: D1 API returned: (400 Bad Request) {\"success\":false,\"errors\":[\"Code: invalid_enum_value ~ Path: action ~ Message: Invalid enum value. Expected 'query' | 'execute' | '_tracing', received 'metadata'\"]}"
    }
  ],
  "messages": []
}

{
  "result": null,
  "success": false,
  "errors": [
    {
      "code": 7501,
      "message": "Internal error: D1 API returned: (400 Bad Request) {\"success\":false,\"errors\":[\"Code: invalid_enum_value ~ Path: action ~ Message: Invalid enum value. Expected 'query' | 'execute' | '_tracing', received 'metadata'\"]}"
    }
  ],
  "messages": []
}

{
  "result": null,
  "success": false,
  "errors": [
    {
      "code": 7501,
      "message": "Internal error: D1 API returned: (400 Bad Request) {\"success\":false,\"errors\":[\"Code: invalid_enum_value ~ Path: action ~ Message: Invalid enum value. Expected 'query' | 'execute' | '_tracing', received 'metadata'\"]}"
    }
  ],
  "messages": []
}

{
  "result": null,
  "success": false,
  "errors": [
    {
      "code": 7501,
      "message": "Internal error: D1 API returned: (400 Bad Request) {\"success\":false,\"errors\":[\"Code: invalid_enum_value ~ Path: action ~ Message: Invalid enum value. Expected 'query' | 'execute' | '_tracing', received 'metadata'\"]}"
    }
  ],
  "messages": []
}

A

aleOP•2/9/24, 8:31 AM

thank you @Matthijs

M

Matthijs•2/9/24, 8:32 AM

welcome, please let us know once you have an eta on a fix if possible

Aale hi, hi 👋 Thank you for reporting these potential issues (and @HardlyWorkin' ...

D

Denver James Duran•2/9/24, 8:39 AM

The log stream also is not working

S

Softbear•2/9/24, 8:42 AM

in the dashboard I can't access my databases

SSoftbear in the dashboard I can't access my databases

S

Softbear•2/9/24, 8:42 AM

May need to update the status page ...

A

aleOP•2/9/24, 8:49 AM

status page update is on its way

A

aleOP•2/9/24, 8:51 AM

https://www.cloudflarestatus.com/incidents/083d43qlrdzm

D1 users may be experiencing a higher error rate and API latency

S

Softbear•2/9/24, 8:57 AM

seems resolved for me

TTzewat I am facing the same problem. I wonder if this affects some users or the whole ...

Ɽ

ⱤØ₥₳₦•2/9/24, 9:04 AM

Same here in India

M

Matthijs•2/9/24, 9:06 AM

fixed for me. thanks for the quick response @geelen and @ale

ⱤⱤØ₥₳₦I just started using it, just 5 rows db (tutorial sql) still takes 500 ms this i...

Ɽ

ⱤØ₥₳₦•2/9/24, 9:06 AM

Jeez so I think this D1 ongoing issue caused this

T

Tzewat•2/9/24, 9:07 AM

d1 works now, thx for the team!

ⱤⱤØ₥₳₦Same here in India

Ɽ

ⱤØ₥₳₦•2/9/24, 9:11 AM

Fixed

ⱤⱤØ₥₳₦I just started using it, just 5 rows db (tutorial sql) still takes 500 ms this i...

M

Max (@rozenmd)•2/9/24, 9:13 AM

what does https://where.durableobjects.live/ show for you?

Where Durable Objects Live

Tracking where Durable Objects are created, wherever you are in the world.

MMax (@rozenmd)what does https://where.durableobjects.live/ show for you?

Ɽ

ⱤØ₥₳₦•2/9/24, 9:16 AM

MMax (@rozenmd)what does https://where.durableobjects.live/ show for you?

Ɽ

ⱤØ₥₳₦•2/9/24, 9:16 AM

AMD, Ahmedabad, India

AAdi Question, I see with the latest workerd release on github the bugfix for row wri...

M

Max (@rozenmd)•2/9/24, 9:16 AM

workerd releases tend to go live within a week or so

ⱤⱤØ₥₳₦AMD, Ahmedabad, India

M

Max (@rozenmd)•2/9/24, 9:18 AM

so your nearest db is unfortunately 200ms away - the queries themselves are fast (if you run

npx wrangler d1 insights <dbname>

npx wrangler d1 insights <dbname>

you can see query timing data)

MMax (@rozenmd)so your nearest db is unfortunately 200ms away - the queries themselves are fast...

Ɽ

ⱤØ₥₳₦•2/9/24, 9:18 AM

It is actually city i live in

MMax (@rozenmd)so your nearest db is unfortunately 200ms away - the queries themselves are fast...

Ɽ

ⱤØ₥₳₦•2/9/24, 9:25 AM

Lol, insanely fast, but not with worker

ⱤⱤØ₥₳₦Lol, insanely fast, but not with worker

M

Max (@rozenmd)•2/9/24, 9:26 AM

like I said, your nearest DB is 200ms away

MMax (@rozenmd)like I said, your nearest DB is 200ms away

Ɽ

ⱤØ₥₳₦•2/9/24, 9:28 AM

How? DB is less than 10 km far

ⱤⱤØ₥₳₦How? DB is less than 10 km far

M

Max (@rozenmd)•2/9/24, 9:29 AM

a worker in AMD created a durable object in Singapore

M

Max (@rozenmd)•2/9/24, 9:29 AM

your worker is less than 10km away, but the DB is in SG

MMax (@rozenmd)your worker is less than 10km away, but the DB is in SG

Ɽ

ⱤØ₥₳₦•2/9/24, 9:30 AM

ohh damn, I got it now

M

Max (@rozenmd)•2/9/24, 9:30 AM

if you add caching to your Worker so it doesn't have to wait for the request from singapore to come back each time, it'll be insanely fast, see: https://developers.cloudflare.com/workers/examples/cache-api/

MMax (@rozenmd)if you add caching to your Worker so it doesn't have to wait for the request fro...

Ɽ

ⱤØ₥₳₦•2/9/24, 9:43 AM

If in the worker there are multiple blocks of code using DB again and again then what will happen?

Each block of code using DB will add Amd to Sin round trip latency to the response time?

ⱤⱤØ₥₳₦If in the worker there are multiple blocks of code using DB again and again then...

M

Max (@rozenmd)•2/9/24, 9:45 AM

If you enable Smart Placement in the dashboard for your Worker, the Worker will move itself to Sin, so only the first hop is slow, and the round-trips will be fast

AAdi Question, I see with the latest workerd release on github the bugfix for row wri...

M

Max (@rozenmd)•2/9/24, 12:21 PM

forgot to mention - once this fix is live, it'll only be applied to new queries, so

wrangler d1 insights

wrangler d1 insights

for historical data (like 31 days) will be inaccurate for a bit

A

Arturo•2/9/24, 12:22 PM

I'm trying to understand this metric, on the selected chart I have "Rows Metrics" selected, as you can see there was some spikes very weird on selected dates, I don't think what I'm doing requires that many rows read, like the max I probably have is like 50K rows, is this an indication of a compromised db? How can I dig deeper?

AArturo I'm trying to understand this metric, on the selected chart I have "Rows Metrics...

M

Max (@rozenmd)•2/9/24, 12:23 PM

run

npx wrangler@latest d1 insights <dbname> --sortBy reads --timePeriod 7d

npx wrangler@latest d1 insights <dbname> --sortBy reads --timePeriod 7d

, will tell you the queries doing this

M

Max (@rozenmd)•2/9/24, 12:25 PM

looking at your tables list counts as a read on every row in every table, so it could be as simple as that (we plan to address this)

A

Arturo•2/9/24, 12:27 PM

the query looks good, the top one is something I'm doing but it looks just too odd

A

Arturo•2/9/24, 12:28 PM

is that counting towards my quota or it's just the queries / write?

A

Arturo•2/9/24, 12:29 PM

I got this metrics from the wrangler:

 "avgRowsRead": 92284,
    "totalRowsRead": 7659605,
    "avgRowsWritten": 0,
    "totalRowsWritten": 0,
    "avgDurationMs": 26.3779313253012,
    "totalDurationMs": 2189.3682999999996,
    "numberOfTimesRun": 83

 "avgRowsRead": 92284,
    "totalRowsRead": 7659605,
    "avgRowsWritten": 0,
    "totalRowsWritten": 0,
    "avgDurationMs": 26.3779313253012,
    "totalDurationMs": 2189.3682999999996,
    "numberOfTimesRun": 83

 "avgRowsRead": 92284,
    "totalRowsRead": 7659605,
    "avgRowsWritten": 0,
    "totalRowsWritten": 0,
    "avgDurationMs": 26.3779313253012,
    "totalDurationMs": 2189.3682999999996,
    "numberOfTimesRun": 83

 "avgRowsRead": 92284,
    "totalRowsRead": 7659605,
    "avgRowsWritten": 0,
    "totalRowsWritten": 0,
    "avgDurationMs": 26.3779313253012,
    "totalDurationMs": 2189.3682999999996,
    "numberOfTimesRun": 83

For the top read spike

AArturo is that counting towards my quota or it's just the queries / write?

M

Max (@rozenmd)•2/9/24, 12:53 PM

it counts towards your quota, but bare in mind you get 25 billion per month with the Workers $5 plan, so you could repeat that week 1470 times before you start getting billed for additional usage

S

Soham•2/9/24, 3:51 PM

Is it possible to access and write data into d1 from python?

S

Soham•2/9/24, 3:51 PM

Is there a direct method or do I have to go through a worker?

M

Max (@rozenmd)•2/9/24, 3:52 PM

going through a worker is the most performant way: https://github.com/elithrar/http-api-d1-example

M

Max (@rozenmd)•2/9/24, 3:53 PM

otherwise there's the API, but it's not running from Workers, and will be slow: https://developers.cloudflare.com/api/operations/cloudflare-d1-query-database

Cloudflare API Documentation

Interact with Cloudflare's products and services via the Cloudflare API

S

Soham•2/9/24, 4:07 PM

Ah nice, I’m not too worried about speed as it’s not a big project

Original message was deleted

M

Max (@rozenmd)•2/10/24, 9:00 AM

I need to improve that error, but D1's beta databases don't allow you to "create" or download your backups yet, you'd need to use a 3rd party tool like https://github.com/nora-soderlund/cloudflare-d1-backups, see also: https://developers.cloudflare.com/d1/reference/time-travel/

Billing data is available via

wrangler d1 insights

wrangler d1 insights

, as well as the Cloudflare dashboard (see the Metrics tab for your D1 database). You can use that data to figure out what D1 will cost you.

V

Vignesh•2/10/24, 11:40 AM

Will exposing database-id expose my database? How to secure D1?

Original message was deleted

V

Vignesh•2/10/24, 11:49 AM

Do I need to implement any security rules? Like how does D1 knows that I am the real owner?

Original message was deleted

V

Vignesh•2/10/24, 11:58 AM

I mean in custom application. Like I just did binding to DB in SvelteKi, so is it enough?

hi, hi 👋 Thank you for reporting these potential issues (and @HardlyWorkin' for raising it). I m

Similar Threads