News for hackers. Stuff that matters.
Re:Unikernels are unfit for production by shin_lao on Friday January 22, 2016 @03:02PM (#10955204)
What it means is that it doesn't matter if they are faster or not because the OS isn't the bottleneck. The bottleneck is the framework or the user application in most of the cases.
Re:Unikernels are unfit for production by dsp1234 on Friday January 22, 2016 @03:11PM (#10955271)
one does not typically find that applications are limited by user-kernel context switches.

the OS isn't the bottleneck.

Curious, then why are we seeing articles here all the time on bypassing the linux kernel for low latency networking?

Re:Unikernels are unfit for production by shin_lao on Friday January 22, 2016 @03:32PM (#10955400)
You don't bypass the kernel, you bypass the TCP stack of the kernel and this is for very specific applications.
Re:Unikernels are unfit for production by elihu on Friday January 22, 2016 @04:58PM (#10955858)
Bypassing the kernel entirely is pretty normal in HPC applications. Infiniband implementations typically memory-map the device's registers into user-space so that applications can send and receive messages without a system call.
Re:Unikernels are unfit for production by shin_lao on Sunday January 24, 2016 @03:50AM (#10961971)
This is not bypassing a kernel. This is called Remote Direct Memory Access (RDMA) and there is still a kernel.

FYI most of the devices inside your computer work through DMA.

Re:Unikernels are unfit for production by EvanPlaice on Friday January 22, 2016 @05:56PM (#10956109)
On in particular is network capturing via libpcap.

It's basically an alternative driver that comes with additional capabilities. Such as capturing promiscuously, and filtering captures.

Re:Unikernels are unfit for production by Twirrim on Friday January 22, 2016 @03:27PM (#10955370)
Just curious, how does a unikernel solve network latency?
Re:Unikernels are unfit for production by masklinn on Friday January 22, 2016 @03:50PM (#10955527)
The issue GP talks about comes from the cost of context-switching on a syscall (going into "kernel mode", performing the call, then going back into "application mode"). There's no context switch in a unikernel.
Re:Unikernels are unfit for production by BrainInAJar on Friday January 22, 2016 @09:16PM (#10956882)
Well, unless you count the hypervisor context switch, which you do.
Re:Unikernels are unfit for production by floatboth on Saturday January 23, 2016 @05:06AM (#10957864)
And if you need super high performance, you can run a unikernel on bare metal.
Re:Unikernels are unfit for production by wmf on Saturday January 23, 2016 @12:09AM (#10957285)
I guess in the glorious future everyone will be using SR-IOV.
Re:Unikernels are unfit for production by silly1 on Saturday January 23, 2016 @05:09AM (#10957874)
Are you assuming SRV-IOV passthrough (which has its own performance profile) ? Because normal virt -definitely- hits a context switch when it goes from unikernels virtual NIC to real NIC, if not twice.
Re:Unikernels are unfit for production by u320 on Friday January 22, 2016 @03:14PM (#10955286)
Unless you are running a load balancer, static http server or in-memory key-value store. Those aren't exactly fringe use cases.
Re:Unikernels are unfit for production by acdha on Friday January 22, 2016 @03:38PM (#10955447)
Even in those examples, it applies mostly to the subset of users who have very large numbers of very simple requests.

The ratio of kernel to userland work is bad if you're receiving an entire network request just to increment a word in memory but usually quite tolerable if, say, you're shoveling video files out and most of the time is spent in something like sendfile() or if your small request actually requires a non-trivial amount of computation.

Re:Unikernels are unfit for production by lmm on Sunday January 24, 2016 @02:24PM (#10963625)
If you're doing nontrivial computation that absolutely dominates the kernel overhead (almost by definition). But how frequent is that? A huge amount of programming boils down to CRUD and very basic transformations, maybe a little bit of fanning out/in, all of which involve minimal actual compute.
Re:Unikernels are unfit for production by shin_lao on Friday January 22, 2016 @03:33PM (#10955412)
Developers of such software might care about unikernels, users, probably less.
Re:Unikernels are unfit for production by Demiurge on Friday January 22, 2016 @03:35PM (#10955419)
This is a correct approach from a developers point of view, however, it is out of scope from the operators point of view. You can always improve your software stack as a developer. These are not exclusive, and are different, should not be conflated.
Re:Unikernels are unfit for production by mwcampbell on Friday January 22, 2016 @03:08PM (#10955246)
That, and you're giving up a lot to gain that marginal increase in performance.