Repository navigation
Does the new SDK expose a pluggable backend abstraction similar to durabletask-go? #757
Description
Activity
Durable Task Scheduler is way too expensive for our usage. We've developed a Cosmos implementation which has been running our orchestrations pretty smoothly so far. We've found a few issues, so it's been getting progressively more robust. We'd happily share the implementation as we'd love to hand it off for others to maintain.
Durable Task Scheduler is way too expensive for our usage. We've developed a Cosmos implementation which has been running our orchestrations pretty smoothly so far. We've found a few issues, so it's been getting progressively more robust. We'd happily share the implementation as we'd love to hand it off for others to maintain.
That's great, thanks for sharing! Would you be open to open-sourcing it? Would love to take a look.
I mainly work with PostgreSQL though, so I'm not too familiar with Cosmos — probably not realistic for me to help maintain it, just curious to see how it's built.It would be great to see support for the self-hosted Durable Task Framework as an additional backend option.
https://github.com/PejmanNik/durabletask-dotnet-aspnet
Reacted by gimmickj
Hi,
I noticed that durabletask-go exposes a clear Backend abstraction with official SQLite and PostgreSQL implementations.
Is the .NET Durable Task SDK intended to support similar pluggable backend implementations in the future, or is the recommended direction to use Durable Task Scheduler for all self-hosted deployments?
I'm trying to understand the long-term extensibility story for self-hosted backends.