Category Archives: Analytics

The engineering of Elastifile

By | | , , , , , , , , , , , , , , , , , , , , , , ,
2 Comments

[Preamble: I was a delegate of Storage Field Day 12. My expenses, travel and accommodation were paid for by GestaltIT, the organizer and I was not obligated to blog or promote the technologies presented in this event]

When it comes to large scale storage capacity requirements with distributed cloud and on-premise capability, object storage is all the rage. Amazon Web Services started the object-based S3 storage service more than a decade ago, and the romance with object storage started.

Today, there are hundreds of object-based storage vendors out there, touting features after features of invincibility. But after researching and reading through many design and architecture papers, I found that many object-based storage technology vendors began to sound the same.

At the back of my mind, object storage is not easy when it comes to most applications integration. Yes, there is a new breed of cloud-based applications with RESTful CRUD API operations to access object storage, but most applications still rely on file systems to access storage for capacity, performance and protection.

These CRUD and CRUD-like APIs are the common semantics of interfacing object storage platforms. But many, many real-world applications do not have the object semantics to interface with storage. They are mostly designed to interface and interact with file systems, and secretly, I believe many application developers and users want a file system interface to storage. It does not matter if the storage is on-premise or in the cloud.

Let’s not kid ourselves. We are most natural when we work with files and folders.

Implementing object storage also denies us the ability to optimally utilize Flash and solid state storage on-premise when the compute is in the cloud. Similarly, when the compute is on-premise and the flash-based object storage is in the cloud, you get a mismatch of performance and availability requirements as well. In the end, there has to be a compromise.

Another “feature” of object storage is its poor ability to handle transactional data. Most of the object storage do not allow modification of data once the object has been created. Putting a NAS front (aka a NAS gateway) does not take away the fact that it is still object-based storage at the very core of the infrastructure, regardless if it is on-premise or in the cloud.

Resiliency, latency and scalability are the greatest challenges when we want to build a true globally distributed storage or data services platform. Object storage can be resilient and it can scale, but it has to compromise performance and latency to be so. And managing object storage will not be as natural as to managing a file system with folders and files.

Enter Elastifile.

Continue reading →

Solid in the Fire

By | | , , , , , , , , , , , , , , , , , , , , , ,
2 Comments

December 22 2015: I kept this blog in draft for 6 months. Now I am releasing it as NetApp acquires Solidfire.

真金不怕紅爐火

The above is an old Chinese adage which means “True Gold fears no Fire“. That is how I would describe my revisited view and assessment of SolidFire, a high performance All-Flash array vendor which is starting to make its presence felt in South Asia.

I first blogged about SolidFire 3 years ago, and I have been following the company closely as more and more All-Flash array players entered the market over the 3 years. Many rode on the hype and momentum of flash storage, and as a result, muddied and convoluted the storage infrastructure market understanding. It seems to me spin marketing ruled the day and users could not make a difference between vendor A and vendor B, and C and D, and so on….

I have been often asked, which is the best All-Flash array today. I have always hesitated to say which is the best because there aren’t much to say, except for 2-3 well entrenched vendors. Pure Storage and EMC XtremIO come to mind but the one that had stayed under the enterprise storage radar was SolidFire, until now.

SolidFire Logo

Continue reading →

The dark ages of data is coming

By | | , , , , , , , , , ,
Leave a comment

A recent report intrigued me. Given the recent uprising of data, data and more data, things are getting a bit absurd about the voluminous data we are collecting and storing. The flip is that we might need all these data for analytics and getting more insight from the data.

The Veritas Darkberg report revealed that a very large percentage of the data collected and stored by organizations are useless data, unknown and unused. I captured a snapshot of the report below:

Screen Shot 2015-11-08 at 8.03.05 AM

From the screenshot above, it shows 54% of the landscape surveyed is dark data, unseen and clogging up the storage. And in an instance, the Darkberg (cross of “Dark” and “Iceberg”) report knocked a lot of sense into this whole data acquisition frenzy we are going through right now.

Continue reading →

MASSive, Impressive, Agile, TEGILE

By | | , , , , , , , , , , , , , , , , , , , , , , , ,
1 Comment

Ah, my first blog after Storage Field Day 6!

It was a fantastic week and I only got to fathom the sensations and effects of the trip after my return from San Jose, California last week. Many thanks to Stephen Foskett (@sfoskett), Tom Hollingsworth (@networkingnerd) and Claire Chaplais (@cchaplais) of Gestalt IT for inviting me over for that wonderful trip 2 weeks’ ago. Tegile was one of the companies I had the privilege to visit and savour.

In a world of utterly confusing messaging about Flash Storage, I was eager to find out what makes Tegile tick at the Storage Field Day session. Yes, I loved Tegile and the campus visit was very nice. I was also very impressed that they have more than 700 customers and over a thousand systems shipped, all within 2 years since they came out of stealth in 2012. However, I was more interested in the essence of Tegile and what makes them stand out.

I have been a long time admirer of ZFS (Zettabyte File System). I have been a practitioner myself and I also studied the file system architecture and data structure some years back, when NetApp and Sun were involved in a lawsuit. A lot of have changed since then and I am very pleased to see Tegile doing great things with ZFS.

Tegile’s architecture is called IntelliFlash. Here’s a look at the overview of the IntelliFlash architecture:

Tegile IntelliFlash Architecture

So, what stands out for Tegile? I deduce that there are 3 important technology components that defines Tegile IntelliFlash ™ Operating System.

  • MASS (Metadata Accelerator Storage System)
  • Media Management
  • Inline Compression and Inline Deduplication

What is MASS? Tegile has patented MASS as an architecture that allows optimized data path to the file system metadata.

Often a typical file system metadata are stored together with the data. This results in a less optimized data access because both the data and metadata are given the same priority. However, Tegile’s MASS writes and stores the filesystem metadata in very high speed, low latency DRAM and Flash SSD. The filesystem metadata probably includes some very fine grained and intimate details about the mapping of blocks and pages to the respective capacity Flash SSDs and the mechanical HDDs. (Note: I made an educated guess here and I would be happy if someone corrected me)

Going a bit deeper, the DRAM in the Tegile hybrid storage array is used as a L1 Read Cache, while Flash SSDs are used as a L2 Read and Write Cache. Tegile takes further consideration that the Flash SSDs used for this caching purpose are different from the denser and higher capacity Flash SSDs used for storing data. These Flash SSDs for caching are obviously the faster, lower latency type of eMLCs and in the future, might be replaced by PCIe Flash optimized by NVMe.

Tegile DRAM-Flash Caching

This approach gives absolute priority, and near-instant access to the filesystem’s metadata, making the Tegile data access incredibly fast and efficient.

Tegile’s Media Management capabilities excite me. This is because it treats every single Flash SSD in the storage array with very precise organization of 3 types of data patterns.

  1. Write caching, which is high I/O is focused on a small segment of the drive
  2. Metadata caching, which has both Read and Write I/O  is targeted to a slight larger segment of the drive
  3. Data is laid out on the rest of the capacity of the drive

Drilling deeper, the write caching (in item 1 above) high I/O writes are targeted at the drive segment’s range which is over-provisioned for greater efficiency and care. At the same time, the garbage collection(GC) of this segment is handled by the respective drive’s controller. This is important because the controller will be performing the GC function without inducing unnecessary latency to the storage array processing cycles, giving further boost to Tegile’s already awesome prowess.

In addition to that, IntelliFlash ™ aligns every block and every page exactly to each segment and each page boundary of the drives. This reduces block and page segmentation, and thereby reduces issues with file locality and free blocks locality. It also automatically adjust its block and page alignments to different drive types and models. Therefore, I believe, it would know how to align itself to a 512-bytes or a 520-bytes sector drives.

The Media Management function also has advanced cell care. The wear-leveling takes on a newer level of advancement where how the efficient organization of blocks and pages to the drives reduces additional and often unnecessary erase and rewrites. Furthermore, the use of Inline Compression and Inline Deduplication also reduces the number of writes to drives media, increasing their longevity.

Tegile Inline Compression and Deduplication

Compression and deduplication are 2 very important technology features in almost all flash arrays. Likewise, these 2 technologies are crucial in the performance of Tegile storage systems. They are both inline i.e – Inline Compression and Inline Deduplication, and therefore both are boosted by the multi-core CPUs as well as the fast DRAM memory.

I don’t have the secret sauce formula of how Tegile designed their inline compression and deduplication. But there’s a very good article of how Tegile viewed their method of data reduction for compression and deduplication. Check out their blog here.

The metadata of data access of each and every customer is probably feeding into their Intellicare, a cloud-based customer care program. Intellicare is another a strong differentiator in Tegile’s offering.

Oh, did I mentioned they are unified storage as well with both SAN and NAS, including SMB 3.0 support?

I left Tegile that afternoon on November 5th feeling happy. I was pleased to catch up with Narayan Venkat, my old friend from NetApp, who is now their Chief Marketing Officer. I was equally pleased to see Tegile advancing ZFS further than the others I have known. With so much technological advancement and more coming, the world is their oyster.

A little yellow elephant

By | | , , , ,
Leave a comment

By now, I believe most of you in the storage networking world would have heard of Hadoop. Hadoop was created by Doug Cutting, while he and his team was working on an open source web search engine called Nutch. The easily recognized little yellow elephant, Hadoop, was Doug Cutting’s son toy, which he made as Hadoop’s mascot. Pretty cool!

And today, Hadoop has become THE platform for Big Data applications. Why?

As I have mentioned before, everything that we do or don’t do, generates data, either as a direct product or in-direct product. I am blogging right now and I am creating data. I was in Singapore the whole of this week and everywhere I go in the MRT stations, I am being watched by the video cameras they have at the station. A new friend in class said that Singapore is the second most “watched” city after London, where there are video cameras mounted everywhere, either discreetly or indiscreetly. And that’s just video data. And there’s plenty of other human activities that generate tons and tons of data.

IDC Digital Universe Report for 2011 said that we have generated 1.8ZB (zettabyte) of data this year alone. I mentioned in my previous blog that this is a gold mine and companies are scrambling to tap on massive amount of data.  Extracting valuable information to anticipate the next trend or predict that next evolution in human preference is akin to the Gold Rush in the wild, wild west in the late 19th century. Folks, Big Data is going to be this generation’s “Digital Gold Rush”.

Sieving, filtering and processing gazillions of data (more unstructured than structured) will not work in defined, well-formatted relational databases. The data model of relational databases will simply break down. And of course, there are different schools of thoughts of different data models, but the Hadoop model seems to be gaining momentum and mind share of data scientists. That is because of Hadoop’s capability to deal with massive unstructured data, processing it and producing results in a small amount of time.

One way to process the pool of massive data is parallel programming. In parallel programming, multi-threading is commonly deployed to achieve the performance and effects of programming. But implementing multi-threading in parallel programming is difficult. Developers often has to deal with LWP (lightweight processes), semaphores, shared memory, mutex (mutually exclusive) locking and so on. Hence this style of programming works with different states on shared data, often resulting in different results in different states, even when using the same programming expression.

Hadoop belongs to another school of programming known as functional programming, where the different states on shared data concept is removed. With that in mind, the dependency on different states is also removed, resulting in a much easier and simpler parallel programming implementation. Hadoop borrows ideas from the MapReduce software framework made well known by Google and the Google File System.

Before, we get to know Hadoop, we must know MapReduce. MapReduce is a framework which allows very large data sets to be processed with a very large set of computer nodes in a cluster. Typically the computational processing is executed in a distributed fashion, spread across many computer nodes and final results are consolidated from the sub-results of these distributed processing nodes.

According to Wikipedia, the 2 key functions of Map Reduce are map() and reduce(). That’s pretty obvious. The extract below was taken from the Wikipedia definition, and explains both functions very well.

“Map” step: The master node takes the input, partitions it up into smaller sub-problems, and distributes them to worker nodes. A worker node may do this again in turn, leading to a multi-level tree structure. The worker node processes the smaller problem, and passes the answer back to its master node.

“Reduce” step: The master node then collects the answers to all the sub-problems and combines them in some way to form the output – the answer to the problem it was originally trying to solve.

The diagram below probably can simplify the concept of MapReduce to the readers.

 

Hadoop is one of the open-source implementations of MapReduce. It is one of the projects of Apache Foundation, and the project has sparked a brand-new niche of data search, data management and data science. The diagram below will allow our readers to juxtapose MapReduce and Hadoop, and comparing them in the simplest fashion.

Hadoop primary development platform is Java. Hadoop’s architecture consists mainly of 2 components – Hadoop Common and a Hadoop-compatible file system, as shown in the diagram below.

Hadoop MapReduce layer above is the file/object access interface to the Hadoop-compatible file system below. HDFS is Hadoop Distributed File System is just one of a few Hadoop-compatible file systems. Other file systems include:

  • Amazon S3 File System as part of the Amazon EC2 Infrastructure-as-a-Service (IaaS) cloud platform
  • CloudStore – a similar Hadoop-like implementation using C++ and also inspired by Google File System
  • FTP file systems
  • HTTP and HTTPS read-only file systems
  • Any file systems accessible with the file:// URL nomenclature

But the main engine of Hadoop is in the MapReduce layer. The 2 core components in this layer is JobTracker and TaskTracker. Both has their own individual roles to play and collectively, they are key cogs in the Hadoop distributed data processing model.

Below are extract I picked up from Wikipedia.

JobTracker submits MapReduce jobs to client applications. The JobTracker pushes work out to available TaskTracker nodes in the cluster, striving to keep the work as close to the data as possible. With a rack-aware filesystem, the JobTracker knows which node contains the data, and which other machines are nearby. If the work cannot be hosted on the actual node where the data resides, priority is given to nodes in the same rack. This reduces network traffic on the main backbone network. If a TaskTracker fails or times out, that part of the job is rescheduled. The TaskTracker on each node spawns off a separate Java Virtual Machine process to prevent the TaskTracker itself from failing if the running job crashes the JVM. A heartbeat is sent from the TaskTracker to the JobTracker every few minutes to check its status. The Job Tracker and TaskTracker status and information is exposed by Jetty and can be viewed from a web browser. Jetty is a Java-based HTTP server, among other things

JobTracker records what it is up to in the filesystem. When a JobTracker starts up, it looks for any such data, so that it can restart work from where it left off.

Scheduling

By default Hadoop uses first-in, first-out (FIFO), and optional 5 scheduling priorities to schedule jobs from a work queue. In version 0.19 the job scheduler was refactored out of the JobTracker, and added the ability to use an alternate scheduler (such as the Fair scheduler or the Capacity scheduler).

Fair scheduler

The fair scheduler was developed by Facebook. The goal of the fair scheduler is to provide fast response times for small jobs and QoS (Quality of Service) for production jobs. The fair scheduler has three basic concepts.

  1. Jobs are grouped into Pools.
  2. Each pool is assigned a guaranteed minimum share.
  3. Excess capacity is split between jobs.

By default jobs that are uncategorized go into a default pool. Pools have to specify the minimum number of map slots, reduce slots, and a limit on the number of running jobs.

Capacity scheduler

The capacity scheduler was developed by Yahoo. The capacity scheduler supports several features which are similar to the fair scheduler.

  • Jobs are submitted into queues.
  • Queues are allocated a fraction of the total resource capacity.
  • Free resources are allocated to queues beyond their total capacity.
  • Within a queue a job with a high level of priority will have access to the queue’s resources.

I took most the extract below from Wikipedia, and I don’t claim to be a knowledgeable person on Hadoop. All the credits go to Wikipedia editors to put Hadoop in layman terms.

Hadoop has certainly won the hearts of the new digital gold rush, Big Data and is slowly becoming a force to be reckoned with among data scientists. Hadoop implementations are powering new frontiers in processing and mining the ever growing data capacity, giving solution providers a simple programming methodology and data model to gain more insights into the vast seas of data and information.

Hadoop has many fans, and slowly becoming the data platform for large companies such as Yahoo!, Facebook, IBM, Amazon, Apple, eBay and many more. Facebook even claims to have the largest Hadoop clusters in the world, growing to 30PB in July of 2011.

This little yellow elephant is going places and one to watch out for.

What should be a Cloud Storage?

By | | , , ,
2 Comments

For us filesystem guys, NAS is the way to go. We are used to store files into network file systems via NFS and CIFS protocols and treating the NAS storage array like a refrigerator – taking stuff out and putting stuff back it. All that is fine and well as long as the data is what I would term as corporate data.

Corporate data is generated by employees, applications and users of the company and for a long time, the power of data creation lies in the hands of the enterprise. That is why storage solutions are designed to address the needs of the enterprise where the data is structured and well defined. How the data is stored; the data is formatted; and how is being accessed are the “boundary” of how the data is being used. Typically a database is used to “restrict” the data created so that the information can be retrieved and modified quickly. And of course, SAN guys will tell you to put these structured data of the data base into their SAN.

For the unstructured data in the enterprise, NAS file systems hold that responsibility. Files such as documents and presentations have a more loosely defined “boundaries”, and hence filesystems are a better natural fit for unstructured data. Filesystems are like a free-for-all container, and able to store and provide access to any files in the enterprise.

But today, as the Web 2.0 applications are already taking over the enterprise, the power of data creation does not necessary lie in the hands of the enterprise applications and users. In fact, it is estimated that the percentage of enterprise data now has exceeded 50% of the enterprise’s total data capacity. With the proliferation of personal devices such as tablets, Blackberries, smart phones, PDAs and so on, individual contributors are generating plenty of data. This situation has been made more acute with Web 2.0 applications, such as Facebook, blogs, social networking, Twitter, Google Search and so on.

Unfortunately, file systems in the NAS category still pretty much the traditional file systems, while the needs of a new type of file system could not be met by the traditional file systems. The paradigm is definitely shifting.  The new unstructured data world needs a new storage concept. I would term this type of storage as “Cloud Storage” because it breaks down the traditional concepts of NAS.

So what basically defined a Cloud Storage? I already mentioned that the type of unstructured data has changed. And the new requirements for unstructured data type  are:

  • The unstructured data type is capable of globally distributed.
  • There will be billions and billions of unstructured data objects created but each object, be it a Twitter tweet, or a uploaded mobile video, or even the clandestine data collected by CarrierIQ, can be accessed easily via a single namespace
  • The storage file system foundation for these new unstructured data type is easily provisioned and management. Look at Facebook. It is easy to setup, get going and the user (and probably the data administrator) can easily manage the user interface and the platform
  • For the service provider of Cloud Storage, the file system must be secure and support multi-tenancy and virtualization of storage resources
  • There should be some form of policy-driven content management. That is why development platforms such as Joomla!, Drupal, WordPress are slowing become enterprise driven to address these unstructured data types.
  • Highly searchable and have a high degree of search optimization. A Google search do have a strong degree of intelligence and relevance to the data being search as well as generating tons of by-product data that feeds the need to understand the consumers or the users better. Hail Big Data!

So when I compare traditional NAS storage solutions such as Netapp or EMC VNX or BlueArc, I ask the question of whether their NAS solutions has these capabilities to meet the requirements of these new unstructured data type.

Most of them, no matter how they package it, is still relying on files as the granular object of storage. And today, most files may have some form of metadata such as file name, owner, size etc, DO NOT, possess the capability of content-aware. Here’s an example when I want to show you:

 

The file properties (part of the file metadata) tell you about the file but little about the content of the file. Today, it requires more than that and the new unstructured data type should look more like this:

If you look at the diagram below, the object on the right (which is the new unstructured data type), display much more information than a typical file in a NAS file system. There additional information becomes the fodder to other applications such as search engines, RSS feeds, robots and spiders and of course, big data analytics.

Here’s another example of what I mean about these extended metadata, and a Cloud Storage storage array is required to work with these new set of parameters and a new set of requirement.

 

There’s a new unstructured data type in town. Traditional NAS systems may not have the right features to work with this new paradigm.

Don’t be white washed by the fancy talk of storage vendors in town. Learn the facts, and find out what is really a Cloud Storage.

It’s time to think differently. It’s time to think of what should be a Cloud Storage.

 

Big data is big headache

By | | ,
1 Comment

IBM claims that we are responsible of for creating 2.5 quintillion bytes of data every day. How much is 1 quintilion?

 

According to the web,

1 quintillion = 1,000,000,000,000,000,000

After billion, it is trillion, then quadrillion, and then quintillion. That’s what 1 quintillion is, with 18 zeroes!

These data comes from everything from social networking updates, meteorology (weather reports), remote sensing maps (Google Maps, GPS, Geographical Information Systems), photos (Flickr), videos (YouTube), Internet search (Google) and so on. The big data terminology, according to Wikipedia, is data that are too large to be handled and processed by conventional data management tools. This presents a new set of difficulties when it comes to collected these data, storing them and sharing them. Indexing and searching big data would require special technologies to be able to mine and extract valuable information from big data datasets, within an acceptable period of time.

According to Wiki, “Technologies being applied to big data include massively parallel processing (MPP) databases, datamining grids, distributed file systems, distributed databases, cloud computing platforms, the Internet, and scalable storage systems.” That is why EMC has paid big money to acquire GreenPlum and IBM acquired Netezza. Traditional data warehousing players such Teradata, Oracle and Ingres are in the picture as well, setting a collision course between the storage and infrastructure companies and the data warehousing solutions companies.

The 2010 Gartner Magic Quadrant has seen non-traditional players such as IBM/Netezza and EMC/Greenplum, in its leaders quadrant.

 

And the key word that is already on everyone’s lips is “ANALYTICS“.

The ability to extract valuable information that helps determines what the next future trend is and personalized profiling will be something that may already arrived as companies are clamouring to get more and more out of our personalities so that they can sell you more of their wares.

Meteorological organizations are using big data analytics to find out about weather patterns and climate change. Space exploration becomes more acute and precise from the tons and tons of data collected from space explorations. Big data analytics are also helping pharmaceutical companies develop new biological and pharmaceutical breakthroughs. And the list goes on.

I am a new stranger into big data and I do not proclaim to know a lot. But terms such as scale-out NAS, distributed file systems, grid computing, massively parallel processing are certainly bringing the data storage world into a new frontier, and it is something we as storage professionals have to adapt to. I am eager to learn and know more about big data. It is a big headache but change is inevitable.