3 ms·
Last time I checked, Firecracker didn't have a very compelling I/O story, which made it in my opinion not completely adequate for running Postgres (or any other
by ahachete 2y ago
Last time I checked, Firecracker didn't have a very compelling I/O story, which made it in my opinion not completely adequate for running Postgres (or any other database).
In contrast, other similar VMM seem to have a better one, like Cloud Hypervisor [1]. Why then FC and not CH? (I've nothing against FC, actually love it and have been using it, but it appears not being the best I/O wise).
[1]: https://github.com/cloud-hypervisor/cloud-hypervisor https://github.com/cloud-hypervisor/cloud-hypervisor
- aayushshah15 2y ago> Firecracker didn't have a very compelling I/O story Can you provide any sources for this claim? We're running Firecracker in production over at blacksmith dot sh and haven't been able to reproduce any perf regressions in Firecracker over CH in our internal benchmarking.
- eyberg 2y agoThe major tradeoff with firecracker is a reduction in runtime performance for a quick boot time (if you actually need that - this obviously doesn't work if your app takes seconds to boot). There are quite a lot of other tradeoffs too like 'no gpu' because that needs some of the support that they remove to make things boot fast. That's why projects like 'cloud hypervisor' exist.
- ahachete 2y agohttps://github.com/firecracker-microvm/firecracker/issues/1600 https://github.com/firecracker-microvm/firecracker/issues/16...
- aayushshah15 2y agoThat issue is from 2020 and has already been addressed. The fact that io_uring support is not "GA" is mostly a semantic aspect that doesn't affect most use-cases.