The problem with performance testing …

May 26, 2015

Had a really good experience with a customer recently, but it highlighted the problems with performance testing, especially using iometer. Now, we use iometer a lot, and it’s a great tool to drill down into a specific set of performance characteristics to show a specific response from a storage system.

However … the problem with such a situation is getting the parameters right so that you are testing the right parameters that match your data.

So this customer was looking at maxcache – our SSD caching functionality that uses SSD drives attached to the 81605ZQ controller to add read and write caching to an array.

Testing with iometer didn’t show that much of an improvement (at least according to the customer). Discussion regarding the test parameters and how long to run a test for (1 minute won’t cut the mustard) saw a big improvement over their original testing (and yes, these guys know what they are doing with their systems so I’m not having a go at any individual system builder here).

So after much testing, it was decided to put the machine into test with real-world customers in a virtual desktop environment (believe is was openE running a whole stack of virtual desktops). Guess what – customers (end users) were as happy as pigs in …

Turns out the real world data is perfectly suited to caching (as suspected by the system builder), but that iometer was not able to accurately reflect the data characteristics of the real-world server. End result: everyone (system builder, datacenter operator, end users) – all happy and amazed at the performance of the system.

So where is the moral in this story? Simply that it’s difficult to play with a test software and come up with something that will closely match the end result of a server used in the real world. Is there an answer to this? Probably not, but I’m suggesting that everyone take performance testing software and the results they get with a grain of salt, and look at testing in the real world, or at least a close simulation.

The results can be very surprising.

Ciao
N

facebooktwitterlinkedinmail

A technical issue regarding RAID build/rebuild

May 13, 2015

Been getting a few questions regarding building RAID arrays recently, and thought it warranted putting something down on paper.

Now I’m talking about RAID 5/6 and other redundant arrays (not RAID 0 – that’s not for “real” data imho). So the questions arise about whether it is possible to restart a server during a raid build or rebuild, and what happens when a drive fails during that process. So let’s take a look at exactly what our controllers actually do in these situations.

RAID Build

If you are building a redundant array using either the clear or build/verify method, then yes, you can power down the server (or the power can go out by any other means) and it won’t hurt your process. We continue building from where we left off, so if the build process gets to 50% and you need to reboot your server, then no worries, it will just continue to build from where it left off – it does not go back to the start again.

If a drive fails during the build process of say a RAID 5, then no worries, the build will continue. When it’s finished, the array will be in degraded mode, and you’ll have to replace the drive, but that’s the normal process. Again, even if a drive fails, you can power down and restart the server during the build process and it will resume building from the point it left off, still finishing in a degraded array that needs fixing.

And if you think drives don’t fail during RAID builds, then think again … I’ve had it happen more than once.

RAID Rebuild

What happens when a drive fails during a rebuild is a bit dependent on the RAID type. Let’s take the example of a RAID 5. A drive fails, so you replace it and the controller starts a rebuild. During that process, another drive fails. You are toast. There is not enough data left for us to calculate the missing data from the first drive failure because now you are missing 2 drives in a RAID 5 and that’s fatal. You need to fix your drives, build a new array and restore from backup.

In a RAID 6 environment it’s slightly different. RAID 6 can support 2 drives failures at the same time, so if a drive fails, you replace it and start a rebuild and then if another drive fails, no worries. The controller will continue to rebuild the array but it will be impacted when finished because it’s still one drive short of a picnic. However the data will be safe during this process, and you’ll just have to replace the second failed drive and let the array rebuild to completion.

Of course, like any of the above, you can power down and restart the server at any time during any of these processes and things will just continue on from where they left off.

Hope that answers a few questions.

Ciao
N

facebooktwitterlinkedinmail

Learning about storage (the hard or easy way) …

May 13, 2015

Google and youtube are wonderful places to get information, but as always there is a question-mark over the authenticity, quality and downright accuracy of the information provided by all and sundry in their posts and blogs. Now while I’m not casting aspersions over those that provide all this wonderful information, wouldn’t it be nice to get something directly from the horses mouth? (Australian-speak for “Vendor”).

Well, you can.

On our website home page (www.adaptec.com) , if you look closely enough, you’ll find the following:

  1. youtube link – some older stuff about our products from the product marketing and an ongoing effort by our Alabama Slammer (you’ll get it when you listen to the video) on technical aspects of how to do stuff with our cards – Liz is by far the best RAID Support Tech in the business so she’s well worth listening to. My only problem with youtube is getting distracted. All that lovely interesting stuff that appears down the right side always looks more interesting than RAID :-)
  2. Facebook – hmmm, I thought this was only for putting pictures of your last meal … however I liked it (pardon the pun) because I found one of my videos on there (you never know where this stuff will end up).
  3. Adaptec University. Last, but certainly not least, this is a major source of all sorts of information on RAID and storage in general, and our products and how to use them in particular. I should know … I spend quite a bit of time writing this stuff (then having it cleaned up by a lovely lady whose English is a whole lot betterer than mine :-)).Yes, you have to register, but no we don’t ask for your first-born as a down-payment (in fact it’s free), but once you are in, there is a wealth of information to peruse through at your leisure. Look at the catelog to see what is available, then go to “my learning” to see what you have completed, not read or not even started yet, etc. You can come back to this as often (or as little) as you like.

So should you stop using Google? Heck no, there is tons of valuable stuff out there (I use it all the time), but you should also consider getting the right word from the right people – Adaptec by PMC.

Ciao
N

facebooktwitterlinkedinmail

The problem with Indonesia …

May 10, 2015

is that … there is no problem with Indonesia!

Recently I spent a week there with our new distributors, where we presented at a Datacenter/Cloud Forum and talked to customers regarding the suitability of our products for their markeplace. Indonesia is a booming economy and the IT sector is growing at a good steady rate, so I can see us doing good business there over the next few years – which means I need to be there on a regular basis.

Since this was my first trip to Jakarta, I thought I’d analyse the problems I found:

1. The people are fantastically friendly and helpful … so no problem there

2. The food is great … so no problem there

3. The customers are smart, up to date and right on top of their game … so no problem there

4. The weather … damn it was humid … so if that’ the only problem I came across I guess I can live with it.

The best part about Indonesia was that I managed to get two weeks holiday after the trip, which was spent chasing my boy around at our BMX National Championships (2nd, Elite Men so a pretty good result), then 4 days on an Island off the coast of Queensland doing some 4-wheel-driving in the sand and relaxing on the beach (with the trip book-ended by some lengthy road trips through outback NSW for good measure).

Hopefully this explains the lack of blog postings over the last month … but now we’re back in business so after I catch up on 4000 emails in my inbox, we’ll be back to regular posts.

Ciao
N

facebooktwitterlinkedinmail