Archive for Tech Field Day

I Like Big Files and I Cannot Lie

You other vendors, can’t deny,
When an array walks in with an itty bitty waste [-ed capacity],
And many spindles in your face
You get sprung, want to pull up tough,
‘Cause you notice that storage was stuffed!

Ok… I’ll stop now! I’m just a bit sad and always wanted an excuse to to use that as a post opener! 🙂

There is a certain, quite specific type of customer whose main requirements revolve around the storage of large data sets consisting of thousands to millions of huge files. Think media / TV / movie companies, video surveillance or even PACS imaging and genomic sequencing. Ultimately we’re talking petabyte-scale capacities – more than your average enterprise needs to worry about!

How you approach storage of this type of data is worlds apart from your average solution!

The Challenges of “Chunky” Data

Typical challenges involve having multiple silos of your data across multiple locations, with different performance and workload characteristics. Then you have different storage protocols for different applications or phases in their data processing and delivery. Each of those silos then requires different skills to manage, and different capacity management regimes.

Sir Mixalot likes big files

On top of that, for the same reason as we moved away from parity groups in arrays to wide striping, these silos then have IO and networking hotspots, wasted capacity (sometimes referred to as trapped white space) and wasted performance, which cannot be shared across multiple systems.

Finally (and arguably most importantly), how do you ensure the integrity, resilience, and durability of this data, as by its very nature, it typically requires long-term retention?

Ideal Solution

What you really need is a single storage system which can not only scale to multi-petabyte capacities with multiple protocols, but is reasonably easy to manage, even with a high admin to capacity ratio.

You then need to ensure that data can also be protected against accidental, or malicious file modification or deletion.

Finally, you need the system to be able to replicate additional copies to remote sites, as backing up petabytes of data is simply unrealistic! Similarly, you may want multiple replicas or additional pools outside of your central repository which all replicate back to the mothership, for example for ROBO or multi-site solutions where editing large files needs to be done locally.

As my good friend Josh De Jong said recently:

Of course, the biggest drawback of using this approach is that you have one giant failure domain. If something somehow manages to proverbially poison your “data lake”, that’s a hell of a lot of data to lose in one go!

DellEMC Isilon

During our recent Tech Field Day 12 session at DellEMC, I was really interested to see how the DellEMC Isilon scale-out NAS system was capable of meeting many of these requirements, especially as this is a product which can trace its heritage all the way back to 2001! In fact, their average customer on Isilon is around 1PB in size, and their largest customer is using 144PB! Scalability, check!

The Isilon team also confirmed that around 70% of their 8,000+ customers trust the solution sufficiently to not use any external backup solution, trusting in SnapshotIQ, SyncIQ and in some cases SmartLock, to protect their data. That’s a pretty significant number!

One thing I am not so keen on with the Isilon (and to be fair, many other “traditional” /  old guard storage vendor offerings) is the complexity and breadth of the licensing; almost all of the interesting features each have to have their own license. If the main benefit to the data lake is simplicity, then I would far rather have a single price with perhaps one or two uplift options for licenses, than an a la carte menu.

In addition, the limit of 50 security domains provides some flexibility for service providers, but then limits the size of your “data lake” to 50 customers. It would be great to see this limit increased in future.Data Lake

The Tekhead Take

Organisations looking to retain data in these quantities need to weigh up the relative risks of using a single system for all storage, versus the costs of and complexity of multiple silos. Ultimately it is down to each individual organisation to work out what closest matches their requirements, but for the convenience of a single large repository of all of your data, the DellEMC Islion still remains a really interesting proposition.

Further Info

You can catch the full Isilon session at the link below:
Dell EMC Presents at Tech Field Day 12

Further Reading

Some of the other TFD delegates had their own takes on the presentation we saw. Check them out here:

Disclaimer: My flights, accommodation, meals, etc at Tech Field Day 12 were provided by Tech Field Day, but there was no expectation or request for me to write about any of the vendors products or services.

Scale-Out. Distributed. Whatever the Name, it’s the Future of Computing

Scale Out

We are currently living in the fastest period of innovation in the technology space which there has probably ever been. New companies spring up every week with new ideas, some good, some bad, some just plain awesome and unexpected!

One of the most common trends I have seen in this however was described in a book I read recently, “The Second Machine Age” by Erik Brynjolfsson & Andrew McAfee. This trend is that the majority of new ideas are (more often than not) unique recombinations of old ones.

Take for example the iPhone. It was not the first smart phone. It was not the first mobile phone, the first touch screen, or the first device to run installable apps. However, Apple recombined an existing set of technologies into a very compelling product.

We also reached a point a while back where clock speeds of CPUs are no longer increasing, and even CPUs are scaling horizontally. Workloads are therefore typically being designed to scale horizontally instead of vertically, taking advantage of the increased compute resources available whilst avoid being locked to vertically scaling clock speeds.

Finally, another trend we have seen in the industry of late is inexpensive and low power CPUs from ARM, being used in all sorts of weird and wonderful places; often providing solutions to problems we didn’t even know we had. Up until now, their place has generally been confined outside of the data centre. I am, however, aware of a number of companies now working on bringing them to the enterprise in a big way!

So, in this context of recombination, imagine then if you could provide a scale-out storage architecture where every single spindle had its own compute directly attached. Then combine many of these “nano-servers” together in a scale-out JBOD form factor on subscription pricing, all managed from a Meraki-style cloud portal… well that’s exactly what Igneous Systems have designed!

Igneous Systems Nano-Servers

One of the coolest things about scaling out like this, is that instead of a small number of large fault domains based around controllers, you actually end up with many tiny fault domains instead. The loss of any one controller or drive is basically negligible within the system and replacements can be sorted at the convenience of the administrators, rather than panicking about replacement of components asap. Igneous claim that you can also scale fairly linearly, avoiding the traditional bottlenecks of a dual controller (or similar) system. It will be interesting to see some performance benchmarks as they become available!

It’s still early days, so they are doing code deployments at some pretty high rates, around every 2 weeks, and to be honest I think there is a bit of work to be done around clarity of their SLAs, but in general it looks like a very interesting platform, particularly when pricing is claimed to be as low as half the price of Amazon S3.

Now as you might expect from a massively distributed solution, the entry point is not small, typically procured in 212TiB chunks, so don’t expect to use it for your SMB home drives! If however you have petabyte-scale data volumes and are looking for an on-prem(ises!) S3 compatible datastore, then its certainly worth looking at Igneous.

The future in the scale-out space is certainly bright, now if only I could get people to refactor their single-threaded applications!… 🙂

Further Info

You can catch the full Igneous session at the link below – it certainly was unexpected and interesting, for sure!

Igneous Systems Presents at Tech Field Day 12

Further Reading

Some of the other TFD delegates had their own takes on the presentation we saw. Check them out here:

Disclaimer: My flights, accommodation, meals, etc at Tech Field Day 12 were provided by Tech Field Day, but there was no expectation or request for me to write about any of the vendors products or services.

Tech Field Day 12 (TFD12) – Preview

Tech Field Day 12 (TFD12)

For those people who haven’t heard of Tech Field Day, it’s an awesome event run by the inimitable Stephen Foskett. The event enables tech vendors and real engineers / architects / bloggers (aka delegates) to sit down and have a conversation about their latest products, along with technology and industry trends.

Ever been reading up on a vendor’s website about their technology and had some questions they didn’t answer? One of the roles of the TFD delegates is to ask the questions which help viewers to understand the technology. If you tune in live, you can also post questions via twitter and the delegates, who will happily ask them on your behalf!

As a delegate it’s an awesome experience as you get to spend several days visiting some of the biggest and newest companies in the industry, nerding out with like-minded individuals, and learning as much from the other delegates as you do from the vendors!

So with this in mind, I am very pleased to say that I will be joining the TFD crew for the third time in San Jose, for Tech Field Day 12, from the 15th-16th of November!

Tech Field Day 12 (TFD12) Vendors

As you can see from the list of vendors, there are some truly awesome sessions coming up! Having previously visited Intel and Cohesity, as well as written about StorageOS, it will be great to catch up with them and find out about their latest innovations. DellEMC are going through some massive changes at the moment, so their session should be fascinating. Finally, I haven’t had the pleasure of visiting rubrik, DriveScale or Igneous to date, so should be very interesting indeed!

That said, if there was one vendor I am probably most looking forward to visiting at Tech Field Day 12, it’s Docker! Container adoption is totally changing the way that developers architect and deploy software, and I speak to customers regularly who are now beginning to implement them in anger. It will definitely be interesting to find out about their latest developments.

If you want to tune in live to the sessions, see the following link:
Tech Field Day 12

If for any reason you can’t make it live, have no fear! All of the videos are posted on YouTube and Vimeo within a day or so of the event.

Finally, if you can’t wait for November, pass the time by catching some of the fun and highlights from the last event I attended:

Storage Field Day 9 – Behind the Curtain

VulcanCast Follow Up – A few thoughts on 60TB SSDs

So last week I was kindly invited to share a ride in Marc Farley‘s car (not as dodgy as it sounds, I promise!).

The premise was to discuss the recent announcements around Seagate’s 60TB SSD, Samsung’s 30TB SSD, their potential use cases, and how on earth we can protect the quantities of data which will end up on these monster drives?!

Performance

As we dug into a little in the VulcanCast, many use cases will present themselves for drives of this type, but the biggest challenge is that the IOPS density of the drives not actually very high. On a 60TB drive with 150,000 read IOPS (and my guess but not confirmed is ~100,000 or fewer write IOPS), the average IOPS per GB is actually only a little higher than that of SAS 15K drives. When you start adding deduplication and compression into the mix, if you are able to achieve around 90-150TB per drive, you could easily be looking at IOPS/GB performance approaching smaller 10K SAS devices!

Seagate 60TB SSD Vulcancast flash is fastThe biggest benefit of course if that you achieve this performance in a minuscule footprint by comparison to any current spindle type. Power draw is orders of magnitude lower than 10/15K, and at least (by my estimates) at least 4x lower than using NL-SAS / SATA at peak, and way more at idle. As such, a chunk of the additional cost of using flash for secondary tier workloads, could be soaked up by your space and power savings, especially in high-density environments.

In addition, the consistency of the latency will open up some interesting additional options…

SAS bus speeds could also end up being a challenge. Modern storage arrays often utilise 12GB SAS to interconnect the shelves and disks, which gives you multiple SAS channels over which to transfer data. With over half a PB of usable storage in just a dozen drives, which could be 1PB with compression and dedupe, and that’s a lot of storage to stick on a single channel! In the long term, faster connectivity methods such as NVMe will help, but in the short-term we may even have to see some interesting scenarios with one controller (and channel) for every few drives, just to ensure we don’t saturate bandwidth too easily.

Seagate 60TB SSD Vulcancast Big DataUse Cases

For me, the biggest use cases for this type of drive are going to be secondary storage workloads which require low(ish) latency, a reasonable number of predominantly Read IOPS, and consistent performance even when a little bit bursty. For example:

  • Unstructured data stores, such as file / NAS services where you may access data infrequently, possibly tiered with some faster flash for cache and big write bursts.
  • Media storage for photo and video sites (e.g. facebook, but there are plenty of smaller ones such as Flickr, Photobox, Funky Pigeon, Snapfish, etc. Indeed the same types of organisations we discussed at the Storage Field Day roundtable session on high performance object storage. Obviously one big disadvantage here, would be the inability to dedupe / compress very much as you typically can’t expect high ratios for media content, which then has the effect of pushing up the cost per usable GB.
  • Edge cache nodes for large media streaming services such as NetFlix where maximising capacity and performance in a small footprint to go in other providers data centres is pretty important,whilst being able to provide a consistent performance for many random read requests.

For very large storage use cases, I could easily see these drives replacing 10K and if the price can be brought down sufficiently, for highly dedupable (is that a word?) data types, starting to edge into competing with NL SAS / SATA in a few years.

Data Protection

Here’s where things start to get a little tricky… we are now talking about protecting data at such massive quantities, failure of just two drives within a short period, has the potential to cause the loss of many hundreds of terabytes of data. At the same time, adding additional drives for protection (at tens of thousands of dollars each) comes with a pretty hefty price tag!

Seagate 60TB SSD Vulcancast data protectionUnless you are buying a significant number of drives, the cost of your “N+1”, RAID, erasure coding, etc is going to be so exorbitant, you may as well buy a larger number of small drives so you don’t waste all of that extra capacity. As such, I can’t see many people using these drives in quantities of less than 12-24 per device (or perhaps per RAIN set in a hyper-converged platform), which means even with a conservatively guestimated cost of $30k per drive, you’re looking at the best part of $350-$700k for your disks alone!

Let’s imagine then, the scenario where you have a single failed drive, and 60TB of your data is now hanging in the balance. Would you want to replace that drive in a RAID set, and based on the write rates suggested so far, wait 18-24 hours for it to resync? I would be pretty nervous to do that myself…

In addition, we need to consider the rate of change of the data. Let’s say our datastore consists of 12x60TB drives. We probably have about 550TB or more of usable capacity. Even with a rate of change of just 5%, we need to be capable of backing up 27TB from that single datastore per night just to keep up with the incrementals! If we were to use a traditional backup solution against something like this, to achieve this in a typical 10-hour backup window will generate a consistent 6Gbps, never mind any full backups!

Ok, let’s say we can achieve these kinds of backup rates comfortably. Fine. Now, what happens if we had failure of a shelf, parity group or pool of disks? We’ve probably just lost 250+TB of data (excluding compression or dedupe) which we now need to restore from backup. Unless you are comfortable with an RTO measured in days to weeks, you might find that the restore time for this, even over a 10Gbps network, is not going to meet your business requirements!!!

This leaves us with a conundrum of wondering how we increase the durability of the data against disk failures, and how do we minimise the rebuild time in the event of data media failure, whilst still keeping costs reasonably low.

Seagate 60TB SSD VulcancastToday, the best option seems to me to be the use of Erasure Coding. In the event of the loss of a drive, the data is then automatically rebuilt and redistributed across many or all of the remaining drives within the storage device. Even with say 12-24 drives in a “small” system, this would mean data being rebuilt back up to full protection in 30-60 minutes, instead of 18-24 hours! That said, this assumes the connectivity on the array bus / backplane is capable of handling the kinds of bandwidth generated by the rebuilds, and that this doesn’t have a massive adverse impact on the array processors!

The use of “instant restore” technologies, where you can mount data direct from the backup media to get up and running asap, then move the data transparently in the background also seems to me to be a reasonable mitigation. In order to maintain a decent level of performance, this will likely also drive the use of flash more in the data protection storage tiers as well as production.

The Tekhead Take

Whatever happens, the massive quantities of data we are beginning to see, and the drives we plan to store them on are going to need to lead us to new (as yet, not even invented) forms of data protection. We simply can’t keep up with the rates of growth without them!

VulcanCast

Catch the video here:

The video and full transcript are also available here:
Huge SSDs will force changes to data protection strategies – with @alexgalbraith

%d bloggers like this: