Is this a free account, or do you pay for Workers btw?
Is this a free account, or do you pay for Workers btw?
M(trying to work out if this "exceeded maximum DB size" error is because your account-wide limit is being hit)
Esince d1 is built ontop of DOs does that mean not all colos are able to have D1 dbs?
Csame regions and locs as DOs
Eis there no way for me to get cloudflare to enable DOs specifically for me in certain colos?
Cmaybe if you pay them a few million in an Enterprise contract
Cthe reason why they're limited is because they need specific regional setups, each write has to be replicated to nearby dcs and confirmed, disaster recovery and such
Eif its a few million for me personally i might as well make my own network
Cidk which region you're looking for but they talked about adding DOs to South America and India, a while ago said they were aiming for end of year but we'll see
Ei personally think that should just be opt-in or you as a consumer should do it urself
ENorth Africa and Middle East (MENA
Cwhat would be opt in?
Eeach write has to be replicated to nearby dcs and confirmed, disaster recovery and suchCCF takes colos offline all the time for maintenance and such, plus just general durability
E:pepethink:
Cdon't think offering a degraded/even more unreliable service in specific locations is a winning path
Eif thats the case why even say its avaiable in the first place
Cwydm?
Eif the servers are unreliable. Why even have them available
CI'm talking about if they were to hypothetically offer D1 in non-supported regions by forgoing replication/all the other properties
Cthere is such thing as storage-less DOs in any colo we've heard about before, I forget the exact details
Cnot sure they could even do north africa, need a bunch of nearby colos/a grouping for reliability and such. South Africa and the Middle east would be cool though, hopefuly one day. When they spun up Oceanic region I got hopeful for more regions but none more yet
Ethey have 4 colos here in Algeria
Ethey have a lot of colos in Middle East
CI think they would need more then that. Look at Oceanic for example: https://where.durableobjects.live/region/oc, 4 PoPs and a total of ~9 colos or so. No idea what their exact requirements are but from the current regions, all got at least 4 PoPs with at least a few being multi-colo. It'd be cool for them to have every region covered though, one can dream
Monly if you want to lose data
Eim willing to manually back it up
MKenton explains it well here: https://blog.cloudflare.com/sqlite-in-durable-objects/#under-the-hood-storage-relay-service
Ethe issue for me is that D1 (DO) isnt ava for me in the countries/cities i need them to be in
Eif they where i'd make my own read replica solution
Mwe're working on it
EI know and i appreciate it. My concern isnt read replicas
Mno i mean running D1/DOs in more places, it'll get there, like australia
Ejust out of curiosity what are the requirements for you guys to have D1/DOs in places
Mhas to be close enough/fast enough to a certain number of colos (forgot how many)
DI'm pretty sure some of us have already said that storageless DOs would be nice to have in more locations for websocket gateways 
Ethis would also enable us to make read replicas
Eread-replicas disappearing isnt an issue so yeah
JDynamodb cut pricing in half. Time for Cloudflare to do the same. 
VAny updates on if D1 will get jurisdictional restrictions? https://community.cloudflare.com/t/d1-location-hints-explanation-and-improvement/634791
JIs there any montioring of what queries are reading the most rows? 
MIs there a way to set upper limits on how many read/write rows per operation?
EIssues on d1 continue. New issues also coming: Cant read from request stream after response has been sent.
Internal error while starting up D1 DB storage caused object to be reset.
Network connection lost.
Etc…
Its been a month now that we receive this issues constantly .
Internal error while starting up D1 DB storage caused object to be reset.
Network connection lost.
Etc…
Its been a month now that we receive this issues constantly .
- Cant read from request stream after response has been sent.* started yesterday, and it correlates with the time that the latency issue was identified : https://www.cloudflarestatus.com/incidents/919j0tc3dmmd
MWhat's your databaseId?
Mand you haven't redeployed your Worker since yesterday?
ENo, we have stopped pushes for 3 weeks now to keep this issue tracked, so we don’t get sidetracked on where the issue is coming from
MI ask because
Cant read from request stream after response has been sent is typically a Worker issueENo pushes
EAlso errors like max size reached are coming in..
VYes, we're planning for H1 2025
