If all I need is access to an sqlite database for my app that does support transaction, do durable o

If all I need is access to an sqlite database for my app that does support transaction, do durable objects provide that capability at this time?

Cctrl-z If all I need is access to an sqlite database for my app that does support trans...

lambrospetrou•6/13/25, 9:23 AM

Yes, you can use SQLite DOs directly

SSilvan the three dots on the right and then explore data

workerBee•6/13/25, 11:44 AM

Thanks - I was waiting for that too... I wish the default table sort was ID DESC, but otherwise, this is a great update!

Larry•6/13/25, 12:24 PM

Brayden must be a proud papa today

Brayden•6/13/25, 2:33 PM

I couldn't be prouder

Brayden•6/13/25, 2:33 PM

Also we have a new "Explore Data" button coming when you click into a D1 instance, the prod deploy for that I think was moved to Monday so apologies for teasing that one a bit early the other day

beingmudit [WePlayDOS]•6/13/25, 3:17 PM

Ahh i see. No worries, thanks for updating Brayden++++
I guess engineers are handling with a massive incident ATM.

SSuccessful-Arm-3762 Hi cloudflare devs, I have a D1 database that I'm using for my production use-ca...

Successful-Arm-3762•6/13/25, 4:38 PM

bumping this again. Hopefully someone is able to help me.

SSuccessful-Arm-3762 bumping this again. Hopefully someone is able to help me.

lambrospetrou•6/13/25, 5:15 PM

This was okay until I was the primary user, but I'm now looking to having more tenants and convert to a 1 db/tenant architecture, so therefore reaching out for help, as I would like to keep using D1.

I am not sure what's the question here, though.

Regarding the slow latencies, it can be anything. Next time they happen send us the database and account IDs so that we can investigate. When the underlying datacenter has network issues certain operations take longer, so that's just one possible cause. The network can also impact just the response on its way back, not necessarily the query processing. Especially if you return MBs So, without exact timeranges we can't say what is the cause.

Tux•6/13/25, 6:22 PM

Hello guys, I have a question about the pricing of D1 of the paid plan. When it says "First 25 billion / month included" in the Rows read, does that mean I only get 25 billion for the first month and then get charged for the usage or does that reset and I get that every month

TTux Hello guys, I have a question about the pricing of D1 of the paid plan. When it ...

Chaika•6/13/25, 6:25 PM

It means every month the first 25 billion rows read are included with Workers Paid, and you only pay for any over

CChaika It means every month the first 25 billion rows read are included with Workers Pa...

Tux•6/13/25, 6:25 PM

Ah alright perfect. Thanks a lot

SSilvan the three dots on the right and then explore data

gruntlord6•6/13/25, 11:04 PM

I wish it was the default, it's so much nicer then the old UI

CC Tried to use Wrangler 4.19.2, the issue still exists. Unable to export remote Da...

C•6/14/25, 12:11 AM

My database down for 10 hours.
https://github.com/cloudflare/workers-sdk/issues/9578
I can not export, need my data back.

Account ID
32c307ab568d8c7e7f64f6694bce4ee4

Database ID
c932e1d8-5c3f-4c4b-b8d4-cd9352b58a85

GitHub

Export D1 DataBase Error: cell.replace is not a function · Issue #...

What versions & operating system are you using? Windows 11, Node.js & NPM, wrangler 4.16.1 / 4.19.2 / 4.20.0 (All tested) Please provide a link to a minimal reproduction No response Describ...

782-627•6/14/25, 12:13 AM

CC My database down for 10 hours. https://github.com/cloudflare/workers-sdk/issues/...

lambrospetrou•6/14/25, 8:03 AM

I just rebooted your database so the export is killed. The error you are getting means there is a mismatch between the actual data you store and the schema defined in the table. So, the current D1 export logic that creates the SQL statements fails. It's one of the weirdness of the current export implementation and is planned to be reworked.

Llambrospetrou I just rebooted your database so the export is killed. The error you are getting...

lambrospetrou•6/14/25, 8:05 AM

Unfortunately, that error is not transient, it's based on the stored data, so if you try to export again, the same thing will most likely happen. I will raise it with the team to see if someone can work on it soon.

Llambrospetrou Unfortunately, that error is not transient, it's based on the stored data, so if...

C•6/14/25, 10:20 AM

Thank you so much! I will not export again until it get fixed.

CC Thank you so much! I will not export again until it get fixed.

gruntlord6•6/15/25, 5:13 PM

you could just have the worker export the data manually if they aren't able to get it doing and you need it sooner

gruntlord6•6/15/25, 5:13 PM

or copy to another database with the correct schema and export from there

gruntlord6•6/15/25, 5:14 PM

personally I would recommend using a second temporary database for the dump anyway since it locks your database while performing it in the current implementation

Disciple of Caoximeihao•6/16/25, 8:54 AM

When using wrangler v3, is it possible to declare database d1 in cloudflare workers outside of the .toml file?

Disciple of Caoximeihao•6/16/25, 8:55 AM

If it is [vars] I have declared it outside the .toml file using .dev.vars

Disciple of Caoximeihao•6/16/25, 8:55 AM

please someone help me

zegevlier•6/16/25, 8:55 AM

Why do you want to do that?

zegevlier•6/16/25, 8:56 AM

I am not sure I know of a way to do that - but knowing why you want it might lead to a workaround

Zzegevlier I am not sure I know of a way to do that - but knowing why you want it might lea...

Disciple of Caoximeihao•6/16/25, 8:59 AM

When I develop I will use a different database, when I am in production I want to use a different database, I don't want to use --env develop

zegevlier•6/16/25, 9:04 AM

If you don't pass the --remote--remote flag to wrangler when starting up your development environment, it should be using a local database. Would that work for you? Otherwise, I don't know of any way to do that without using --env--env.

Note that you could change your npm run devnpm run dev command to include the --env--env flag, so you don't need to type it every time, if that's the issue.

Tory•6/16/25, 1:47 PM

Is it possible to restore data for a specific table in a D1 database? It looks like we experienced some data loss during the recent outage for a specific table.

TTory Is it possible to restore data for a specific table in a D1 database? It looks l...

lambrospetrou•6/16/25, 3:44 PM

It looks like we experienced some data loss during the recent outage for a specific table.

Can you elaborate on this please? Feel free to DM me with as much details as you can. The recent incident could not lead to data loss for D1 from everything we have seen so far. Only startups were impacted and some writes not going through potentially.

You can restore using the bookmark API but that's for the database as a whole, not individual tables.

Gagan Suie•6/16/25, 7:30 PM

did D1 dashboard just remove "Table" tab? and replace it with "Explore Data"?

astra•6/16/25, 7:30 PM

yeah the update just went live to me just now aswell.. some of my columns are missing unless i explicitly search for it

Gagan Suie•6/16/25, 7:31 PM

freaked me out because i was actively making changes and thought i purged my entire DB accidentally. so i restored and lost some data. T_T

Aastra yeah the update just went live to me just now aswell.. some of my columns are mi...

Brayden•6/16/25, 7:51 PM

Can you elaborate a bit on this one? This doesn't sound like expected behavior.

GGagan Suie freaked me out because i was actively making changes and thought i purged my ent...

gruntlord6•6/16/25, 9:08 PM

you can also restore back to before you restored it

gruntlord6•6/16/25, 9:08 PM

time travel is magical

alphem•6/16/25, 11:29 PM

I'm not sure if this is the place to post, but I am not a fan of the front end updates through the dashboard. The fact that I cannot easily see what tables populate the db is very frustrating. Please bring this back. The user interface had a few issues, but this is not better.

GGagan Suie did D1 dashboard just remove "Table" tab? and replace it with "Explore Data"?

gruntlord6•6/16/25, 11:41 PM

I don't have a tab for that, I have to go to the other menu

gruntlord6•6/16/25, 11:42 PM

not exactly what I meant when I said I preffered the other way, I meant I wanted that to show on the table tab

Aalphem I'm not sure if this is the place to post, but I am not a fan of the front end u...

Visal .In•6/17/25, 12:56 AM

What do you mean you can’t see which table is populating the database? Do you see the “Explore Data” button on the right side?

alphem•6/17/25, 12:57 AM

I see that well enough. My point is that I can't just see everything from the dashboard now. I have to open a new tab in order to see it.

alphem•6/17/25, 12:58 AM

It would be best if "explore data" was just a tab on the dashboard.

Aalphem It would be best if "explore data" was just a tab on the dashboard.

Visal .In•6/17/25, 1:04 AM

Got it. Now I understand what you mean. We’ll continue to improve it.

VVisal .In Got it. Now I understand what you mean. We’ll continue to improve it.

alphem•6/17/25, 3:18 AM

TY. Don't mean to complain overmuch, you guys have an amazing product here.

�

😈 Donkey 💫•6/17/25, 3:45 AM

the table tab has gone, why?

dsds10•6/17/25, 7:10 AM

great job on adding the new tables viewer! that's A LOT better

kd•6/17/25, 8:59 PM

Hi! I'm looking for some clarity on D1 cold starts. I have a ~600MB database in D1 right now which will mostly be used as a read-only DB, fronted by a Worker-based react-router website. I have not rolled it out into production yet, but the cold starts of ~5 seconds after every 5-10 minutes of idling makes wonder if I made a right choice for a low-traffic use case. I mean, if there is only a couple visitors an hour, they will all get these huge delays on some of the interactions where I very much preferred a 50ms response that's happening after

Are there any rough numbers of when the D1 goes to sleep and how much traffic should be there to avoid cold starts?

kd•6/17/25, 9:02 PM

I am querying D1 from separate react-router's "API" routes, which I believe should be bundled altogether in a single worker. The routes that do not touch D1 have like a 50ms ttfb, while routes that query D1 have this super long cold start

kd•6/17/25, 9:03 PM

the worst part is that while I was writing this message (~1-2 minutes), it happened again. The subsequent requests are all 40-50 ms though, so I'd assume it's not the heavy joins or something like this (I almost don't have any heavy sql queries)

Kkd Hi! I'm looking for some clarity on D1 cold starts. I have a ~600MB database in ...

lambrospetrou•6/17/25, 10:42 PM

The databases go to sleep after 1-2 minutes of inactivity (due to the underlying Durable Object hibernation/eviction).
So, in your case you will experience booting up the database all the time if you query once every 5-10 minutes.

Now, in theory, unless the underlying machine your database is hosted is bad all the time, your boots should be hot after the first one, and startups should be faster without needing to fetch from cold storage all the time. If you share your database ID/account ID I can double check to see the cold starts you were having.

Kkd I am querying D1 from separate react-router's "API" routes, which I believe sho...

gruntlord6•6/18/25, 12:26 AM

if you are using react router anyway you can easily add in durable objects if you want faster performance

If all I need is access to an sqlite database for my app that does support transaction, do durable o

Similar Threads

Similar Threads

Similar Threads