TechOpsGuys.com Diggin' technology every day

October 8, 2010

I/O Virtualization for mortals

Filed under: Networking,Virtualization — Nate @ 10:24 pm

This product isn’t that new but haven’t seen many people talk about it, I first came across it a few weeks ago certainly looked very innovative.

I don’t know who started doing I/O virtualization first, maybe it was someone like Xsigo, or maybe it was HP with their VirtualConnect or maybe it was someone else, but the space has heated up in the past couple of years.

Neterion, a name that sounds familiar but I can’t quite place it… a company by the name of Exar may of bought them or something. But anyways there is this interesting virtualized NIC that they have – the X3120 V-NIC looks pretty cool –

Neterion’s family of 10 Gigabit Ethernet adapters offer a unique multi-channel device model. Depending upon the product, a total of between eight and seventeen fully independent, hardware-based transmit and receive paths are available; each path may be prioritized for true Quality-of-Service support.

I/O Virtualization Support

  • Special “multi-function PCI device” mode brings true IOV to any industry-standard server. In multi-function mode, up to 8 physical functions are available (more in ARI-capable systems). Each physical function appears to the system as an independent Ethernet card
  • Unique, hardware-based multi-channel architecture mitigates head-of-line blocking and allows direct data transfer between hardware channels and host-based Virtual Machines without hypervisor intervention (greatly reducing CPU workload)
  • VMware® NetQueue support
  • Dedicated per-VF statistics and interrupts
  • Support for function-level reset (FLR)
  • Fully integrated Layer 2 switching function

I removed some bullet points of things to shorten the entry a bit and those things I wasn’t exactly sure what they did anyways! Anyone know what ARI means above?

Never used the product, but it is very nice to see such a product in the market place, to get Virtual Connect “like” functionality (at least as far as the virtual NICs go, I know theres a lot of other advantages to VC) in your regular rack mount systems from any vendor and at least potentially connect to any 10GbE switch, as far as I can tell there’s no special requirements for a specific type of switch.

How inefficient can you get?

Filed under: Storage — Tags: , , , — Nate @ 8:29 pm

[ the page says the system was tested in Jan 2010, so not recent, but I don’t recall seeing it on the site before now, in any case it’s still crazy]

I was about to put my laptop down when I decided hey let’s go over to SPEC and see if there are any new NFS results posted.

So I did, you know me I am into that sort of stuff. I’m not a fan of NFS but for some reason the SPEC results still interest me.

So I go and see that NEC has posted some results. NEC isn’t a very well known server or even IT supplier in the U.S. at least as far as I know. I’m sure they got decent market share over in Asia or something.

But anyways they posted some results, and I have to say I’m shocked. Either there is a glaring typo or that is just the worst NAS setup on the face of the planet.

It all comes down to usable capacity. I don’t know how you can pull this off but they did – they apparently have 284 300GB disks on the system but only have 6.1 TB of usable space! That is roughly 83TB of raw storage and they only manage to get something like 6% capacity utilization out of the thing?

Why even bother with disks at all if your going to do that? Just go with a few SSDs.

But WAIT! .. WAIT! It gets better. That 6.1 TB of space is spread across — wait for it — 24 file systems.

12 filesystems were created and used per node. One of 24 filesystems consisted of 8 disks which were divided into two 4-disk RAID 1+0 pools, and each of the other 23 filesystems consisted of 12 disks which were divided into two 6-disk RAID 1+0 pools. There were 6 Disk Array Controllers. One Disk Array Controller controlled 47 disks, and each one of the other 5 controlled 48 disks.

I mean the only thing I can hope for is that the usable capacity is in fact a big typo.

Total Exported Capacity 6226.5GB

But if it’s not I have to hand it to them for being brave enough to post such terrible results. That really takes some guts.

 

Windows mobile: the little OS that couldn’t

Filed under: General — Nate @ 8:15 pm

I don’t know how to feel, a mixture of happiness and sorrow I suppose. Happy that MS is still unable to get any traction in the mobile space, and sad that they have spent so much time and effort and have gotten nowhere, I feel sorry for them, just a little sorry though.

From our best friends at The Register, an article quoting some folks at Gartner –

Gartner said that Windows Phone 7 will provide a fillip to Microsoft’s worldwide mobile market share, pushing it up from 4.7 per cent this year to 5.2 per cent in 2011, but it’s share will fall again to 3.9 per cent by 2014.

That’s just sad, for the worlds biggest software company. I mean I have a Palm pre and I know their market share is in the toilet, not that it really matters at this point since they sold out to HP they made their money. But Palm had a microscopic amount of resources compared to Microsoft. Microsoft has been trying this for I have to think at least a decade now. If the above is true, if by 2014 they have 4% market share, what would you do if you were MS, spending 14 years to get 4% market share?

I never understood the mobile space, and I worked at one of the earliest companies that capitalized on the space earlier this decade, selling ringtones and wallpapers like nobody’s business. All I could think was what are these people thinking buying all this crap. But they just kept coming and coming. Like moths to a flame or something. Nobody was worse than Sprint though before they ended up partnering with that company I was with. I remember back in … 2004? I looked to see what Sprint had to offer as far as ringtones and stuff and they actually wanted you to rent them, that’s right, pay $2.99 or whatever for a ringtone and then have to buy it again in 90 days. That practice stopped after they bought Nextel which was already a customer of ours at the time and Sprint was merged into the service that we provided.

If it wasn’t for Classic, I would still be using a trusty Sanyo feature phone, booted up in about 15 seconds, crashed maybe once or twice a year, worked really well as a phone, and battery lasted a good long time too.

I noticed the battery life on my Palm went through the roof practically after I stopped using the MS Exchange plugin all this time I thought the battery life was crap when it was that stupid plugin draining it.

Looking forward to the PalmPad from HP myself. I won’t go near an Apple product. I don’t trust Google either, so Android is out 🙂

 

New TechOpsGuy: Scott

Filed under: General — Nate @ 7:46 pm

Well what do you know, I make a passive comment about whether or not someone will join up with me in the future for this site in response to Robin’s comments, and not long after I get a volunteer. He sounds really good, focusing more on the software end of things than hardware, but from what I read of him he is similar to me — seeking out best of breed technologies to make his life better.

He seems to be a Linux expert, automation expert, knows everything there is about open source stuff. And even seems to knows networking having built an Extreme-based network recently (sorry couldn’t resist).

So welcome Scott to the site, and I’m sure he will post something introducing him self at some point in the not too distant future.

I hope he knows what he is getting into – the bar is high for publishing content! It does take time to get in the grove, it took me a few months to figure out how to write stuff in this manor.

This site gets far more traffic than I ever really thought, I’m really surprised, impressed, and flattered. I get enormous positive feedback and I really feel this site has done more for my professional career than well any job I have had. I’m really glad it’s here and will do my best to continue posting my thoughts.

Thanks a lot to all of the readers, whether you like what I have to say or not 🙂

I’m getting close to launching my non technical blog, hopefully this weekend. Can you believe that I have so much more to say that I need a second blog? You have probably seen hints of what this new blog will contain from past posts. Those of you who know me, well know what I have to say. I will try to keep it as tame and objective as this site (regardless of how I may come across sometimes I try very hard to be calm and controlled, those that know me know this is true I am RAW and honest — too raw for some I’m sure on occasion), but can’t make promises.

Manually inflating the memory balloon

Filed under: Virtualization — Tags: , — Nate @ 12:10 am

As I’m sure you all know, one of the key technologies that VMware has offered for a long time is memory ballooning to free memory from idle guest OSs in order to return that memory to the pool.

My own real world experience managing hundreds of VMs in VMware has really made me want to do one thing more than anything else:

Manually inflate that damn memory balloon

I don’t want to have to wait until there is real memory pressure on the system to reclaim that memory. I don’t use windows so can’t speak for it there, but Linux is very memory greedy. It will use all the memory it can for disk cache and the like.

What I’d love to see is a daemon (maybe vmware-tools even) run on the system monitoring system load, as well as how much memory is actually used, which many Linux newbies do not know how to calculate, using the amount of memory reported being available by the “free” command or the “top” command is wrong. True memory usage on Linux is best calculated:

  • [Total Memory] – [Free Memory] – [Buffers] – [Cache] = Used memory

I really wish there was an easy way to display that particular stat, because the numbers returned by the stock tools are so misleading. I can’t tell you how many times I’ve had to explain to newbies that just because ‘free’ is saying there is 10MB available that there is PLENTY of ram on the box because there is 10 gigs of memory in cache. They say, “oh no we’re out of memory we will swap soon!”. Wrong answer.

So back to my request. I want a daemon that runs on the system, watches system load, and watches true memory usage, and dynamically inflates that baloon to return that memory to the free pool, before the host runs low on memory. So often VMs that run idle really aren’t doing anything, and when your running on high grade enterprise stoage, well you know there is a lot of fancy caching and wide striping going on there, the storage is really fast! Well it should be. Since the memory is not being used(sitting in cache that is not being used) – inflate that balloon and return it.

There really should be no performance hit. 99% of the time the cache is a read cache, not a write cache, so when you free up the cache the data is just dropped, it doesn’t have to be flushed to disk (you can use the ‘sync’ command in a lot of cases to force a cache flush to see what I mean, typically the command returns instantaneously)

What I’d like even more than that though is to be able to better control how the Linux kernel allocates cache, and how frequently it frees it. I haven’t checked in a little while but last I checked there wasn’t much to control here.

I suppose that may be the next step in the evolution of virtualization – more intelligent operating systems that can be better aware they are operating in a shared environment, and return resources to the pool so others can play with them.

One approach might be to offload all of storage I/O caching to the hypervisor. I suppose this could be similar to using raw devices(bypasses several file system functions). Aggregate that caching at the hypervisor level, more efficient.

 

Powered by WordPress