Live data from Hacker News

Microsoft and Red Hat partner

blogs.microsoft.com

51–60 of 177 posts

Re: Microsoft and Red Hat partner

#51

In the last 5 years we've experimented with Linux on top of hyper-v a few times. The basic I/O, compile & runtime performance was significantly inferior to xen & kvm (didn't bench against vmware, phasing it out due to cost), it wasn't worth any effort to even deploy apps for testing. Therefore I don't see Linux on hyper-v being a compelling option for the cost conscientious technical officer or lead engineer.

I'm getting approximately the same CPU & storage performance out of Azure's A0 machine w/Debian8, as from AWS t2.micro instance with spinning rust storage and Debian8 (both are cheapest offer available, and price is quite similar between them).

Network on Azure though is a totally wrecked thing:

- A0 instance is capped to a miserable 5 megabit/s uplink (downlink is fine);

- If you want TCP/UDP ports open, you need to open them one-by-one, going through a 2-page wizard with 8 fields in total to fill in for each port you are opening, and then waiting 15-20 seconds for each "change" to be applied. Good luck opening a 1024-port range for your SIP server;

- You can't open ICMP/GRE/IPSEC or any other custom protocol;

- There are ton of really weirdly configured network gear between your VM and the Internet - packet drops and private IPs in traceroute, etc...

- On a plus side, you can enable so called "direct server return" and you will get incoming packets with your public IP as a destination IP; a really nice feature sorely missing on AWS.

Re: Microsoft and Red Hat partner

#52

who uses redhat ? (im not being sarcastic)

Anyone who wants to run Oracle on a supported version of Linux, that isn't 'unbreakable'.

If you want support from Oracle and you're not running on RHEL or Oracle's own offering, the answer usually is, "Get back to us when you can reproduce it on a supported flavor of Linux."

Re: Microsoft and Red Hat partner

#54

I think this is just ticking a box for Microsoft since I believe you can get RH on AWS. But let's be honest - how many people will perceive any value in stacking a Microsoft technology (.NET) on top of UNIX? Sure if you already have a .NET app, hosting it in UNIX may give you some benefit. But is anyone really going to write an app from scratch with this in mind? I'm skeptical.

With their release of CoreCLR[1], which combined with the class libraries, I think there is a lot of reason to look at .NET on UNIX. The implementation includes the same JIT and GC that the .NET framework uses, it just cuts out a bunch of bits that likely wouldn't be relevant to people developing server backends (WPF, AppDomains, etc.). For server stuff, .NET is now close to the level of openness and compatibility th…

You may be right, I don't know. I suppose the acid test is when YC companies start posting for front-end developers with C#, .NET and UNIX experience versus what's required today, i.e. either node.js, django, ruby, etc.

Re: Microsoft and Red Hat partner

#57

In the last 5 years we've experimented with Linux on top of hyper-v a few times. The basic I/O, compile & runtime performance was significantly inferior to xen & kvm (didn't bench against vmware, phasing it out due to cost), it wasn't worth any effort to even deploy apps for testing. Therefore I don't see Linux on hyper-v being a compelling option for the cost conscientious technical officer or lead engineer.

But the game isn't the performance on a machine, or even a few machines.

The game are tools like Kubernetes and Apache Mesos. I have X thousand VMs spread out through the world, with a few hundred per datacenter for geo-stability.

If a network link fails, no biggie. Still runs. If a rack dies, still no biggie. Redundancy makes it work. Earthquake/tsunami/natural disaster ? Your stuff still runs, with a bit higher latency and slower throughput.

Re: Microsoft and Red Hat partner

#58

I think this is just ticking a box for Microsoft since I believe you can get RH on AWS. But let's be honest - how many people will perceive any value in stacking a Microsoft technology (.NET) on top of UNIX? Sure if you already have a .NET app, hosting it in UNIX may give you some benefit. But is anyone really going to write an app from scratch with this in mind? I'm skeptical.

I would, .NET and the latest C# is awesome and superior to other frameworks, so yes, the fact that can run in Linux is something very atractive.

C# is a decent language if you like imperative programing. .NET MVC, though, is far from impressive. The routing is crap, Razor is designed to encourage you to dump C# straight into your views, and in general it feels like a "let's do Rails in C#", which I really don't see as the future of web development.

Re: Microsoft and Red Hat partner

#60

I think this is just ticking a box for Microsoft since I believe you can get RH on AWS. But let's be honest - how many people will perceive any value in stacking a Microsoft technology (.NET) on top of UNIX? Sure if you already have a .NET app, hosting it in UNIX may give you some benefit. But is anyone really going to write an app from scratch with this in mind? I'm skeptical.

With their release of CoreCLR[1], which combined with the class libraries, I think there is a lot of reason to look at .NET on UNIX. The implementation includes the same JIT and GC that the .NET framework uses, it just cuts out a bunch of bits that likely wouldn't be relevant to people developing server backends (WPF, AppDomains, etc.). For server stuff, .NET is now close to the level of openness and compatibility th…

I think this is the same problem Google had with Motorola. Sure, Google was playing nice. But the OEMs are afraid how long Google would be playing nice until Google decide to favor it's own phone-making division? Same applies here. How long MS is going to play nice before they decide to favor their own platform. Needless to mention, MS's past history isn't going to help much either.
Post reply on HN