Microsoft Has More SMB3/CIFS Enhancements For Linux 5.16, Including For Performance

[indepe]

That's with regards to "features" of Samba vs KSMBD today. This many equalize down the road a piece. You'd choose Samba over the more primitive KSMBD due to features today.

However it is the result of decisions that were made some time ago, and IO_URING didn't come into being just today. It doesn't mean that it was technically a wise decision to accept the making of an in-kernel implementation in the first place (which then has to be supported forever). And I'm worried that more such decisions will be made.

[cjcox]

I have to? I'd think IO_URING would work for "basic" SMB just as well.

That's with regards to "features" of Samba vs KSMBD today. This many equalize down the road a piece. You'd choose Samba over the more primitive KSMBD due to features today.

[indepe]

With regards to KSMBD, you have to realize that a lot of people want a basic SMB. For anything "real", you need to be using Samba (preferably on something that isn't a Red Hat derivative). And yes, IO_URING can greatly increase the performance of Samba. https://samba.plus/blog/detail/ksmbd...nel-smb-server

I have to? I'd think IO_URING would work for "basic" SMB just as well.

[jacob]

I have 0 background in this but my 2 cents is....
I have been noticing a lot of Windows centric changes to Linux recently to make Linux more compatible with Windows...
By making Linux more compatible with Windows it'll help making the switch from Windows to Linux easier..
So far this year I've seen Ubuntu installation with active directory built in
SMB3 and the Linux kernel
Tons of changes to wine allowing anti cheats work on Linux which let lots of Windows only games work on Linux...

Before people didn't switch to Linux since nothing worked on it soon that won't be an excuse and will allow more people to switch without having to sacrifice functionality

There is certainly lots of cross pollination between Linux and Windows - not just windows centric changes in Linux, it really goes both ways. And I think it's a very good thing. Linux and Windows are by far and large the two most widely used OSes in the world. Compatibility between them is paramount. In fact, being able to develop software that is portable between Linux and Windows is infinitely more important than portability between say Linux and FreeBSD.

[NobodyXu]

Would an IO_URING based SMB server not be better than an in-kernel server? Just as good if not better?

I don't like the apparent recent tendency to do more inside the kernel. I like it much better if technologies like IO_URING improve the ability of userspace to do things extremely efficiently.

I think ksmb supports RDMA:

Remote Direct Memory Access (RDMA) is a technology that allows computers in a network to exchange data in main memory without involving the processor, cache or operating system of either computer. Like locally based Direct Memory Access (DMA), RDMA improves throughput and performance because it frees up resources.

Copied from searchstorage.techtarget.com/definition/Remote-Direct-Memory-Access

[cjcox]

With regards to KSMBD, you have to realize that a lot of people want a basic SMB. For anything "real", you need to be using Samba (preferably on something that isn't a Red Hat derivative). And yes, IO_URING can greatly increase the performance of Samba. https://samba.plus/blog/detail/ksmbd...nel-smb-server

[indepe]

My question above already answered:
Just found this quote on Jens Axboe's twitter (quoting a reference):

"Clearly, those numbers are impressive, but at the same time recent improvements in Samba's IO performance put this into perspective: by leveraging the new “io_uring” Linux API Samba is able to provide roughly 10x the throughput compared to ksmbd."

[cytomax55]

I have 0 background in this but my 2 cents is....
I have been noticing a lot of Windows centric changes to Linux recently to make Linux more compatible with Windows...
By making Linux more compatible with Windows it'll help making the switch from Windows to Linux easier..
So far this year I've seen Ubuntu installation with active directory built in
SMB3 and the Linux kernel
Tons of changes to wine allowing anti cheats work on Linux which let lots of Windows only games work on Linux...

Before people didn't switch to Linux since nothing worked on it soon that won't be an excuse and will allow more people to switch without having to sacrifice functionality

[lacek]

Phoronix: Microsoft Has More SMB3/CIFS Enhancements For Linux 5.16, Including For Performance
(...)in-kernel SMB3 file server...(...)

Why is putting SMB3 file server inside the kernel a good idea? Does it allow something that user-space application would not allow? If there are some bottlenecks that make user appliations slow, perhaps fixing those would allow not only quick SMB3 but other servers as well?

There is 30-70% speedup on the horison. Isn't it because the current implementation does have some features yet, and completing it will also slow it down?

Not really a joke: Why not KHTTPD ? KSSHD? KGNOME? or KBLENDER ? maybe a in-kernel Steam client?

[indepe]

Would an IO_URING based SMB server not be better than an in-kernel server? Just as good if not better?

I don't like the apparent recent tendency to do more inside the kernel. I like it much better if technologies like IO_URING improve the ability of userspace to do things extremely efficiently.