In the following video, Eyal Markovich, Performance Director for Kaminario, talks about how to identify if your database application has a storage I/O bottleneck. He discusses I/O wait time as the root cause of poor database performance and how to assess if you have an I/O wait time problem.
Archive for June, 2011
Does your database have a storage I/O bottleneck?
By Eyal Markovich
Ain’t it Hot, Hot, Hot
SSD IS HOT, BUT DO YOUR HOMEWORK SO YOU DON’T GET BURNED.By Gareth Taube, Vice President Marketing, Kaminario
There’s been a lot of buzz in the SSD market in the past few months, including a public offering, multiple instances of major new funding, and commitments to SSD from the big storage vendors. Fusion-io, maker of Flash memory PCIe server cards, completed its initial public offering on June 9, closing at $22.50 a share and raising $237 million. Kaminario closed a $15 million series C funding round. Violin Memory, maker of flash memory arrays and caching systems, closed a $40 million Series C funding round shortly after raising $35 million in February. Pure Storage raised $28 million in venture capital even though its solution is still in private beta testing and has yet to be revealed. EMC has made a commitment to boosting several of its arrays with flash and of course Oracle bases its Exadata appliance on Flash. Anyone how knows anything about this market is probably already weary of hearing the phrase, “Flash is the new disk and disk is the new tape.”
Identifying your Performance Problem
WHEN IS BAD PERFORMANCE NOT STORAGE I/O?By Eyal Markovich
One of the first posts which I published in the blog was about acceptable I/O latencies. I have recently received a question from a reader related to this topic that I would like to share with you. It a great example to the point I was trying to make in that post.
The reader wrote (I edited the original email to maintain the reader’s privacy):
Hello. I have just read your article about acceptable I/O latencies at http://www.theiostorm.com/whats-an-acceptable-io-latency/ and I think I need your advice.
Apache2 Web server dispatching quite short streamed audios from disk and connecting to remote MySQL. Such machine relies on an Amazon EC2 large instance which has 1 EBS (Elastic Block Device – Disk) for system and another one for data (all the Apache content).
Apparently, the tool munin is reporting quite strangely high DISK LATENCIES
Average of 10.37ms for system disk
Average of 156.22ms for data disk
I am a bit lost because ‘top’ does not actually report a high I/O wait cpu usage ::> Average of 1%
So, do I believe top and seek for the culprit, or other place, or do I believe munin and think My Amazon EBS volumes are guilty of my suffering?
Thanks in advance for your answers and for your article too.
The Complete Package
TAKING THE COMPLEXITY AND COST OUT OF DEPLOYING HIGH-PERFORMANCE STORAGEBy Gareth Taube, Vice President Marketing, Kaminario
In a recent SearchStorage.com blog entitled Sizing up storage performance analysis options, Jeff Boles laments the lack of attention to storage performance in the rush to address capacity and the cloud. “And because of that neglect, many of us are finding we need better performance than we can get from today’s standard approaches,” says Boles. Read the rest of this entry »
Purpose Built for Speed and Safety
KAMINARIO LET’S DRAM SHOW ITS STUFFBy Gareth Taube, Vice President Marketing, Kaminario
The past 10 years have seen major advances in CPU, network, system bus, and memory performance. With its inherent mechanical limitations, however, disk I/O has struggled to keep up. Storage vendors and IT employ clever techniques to push disk I/O to its limits, but in many cases application performance issues still come down to disk I/O bottlenecks.
That’s why it’s puzzling that many vendors have chosen a disk-like architecture for their SSD products. Sure, replacing RAID drives with flash SSD can boost the speed of random reads and, to a lesser extent, writes, but deploying SSD in a legacy disk array is like driving a race car through a busy intersection. As Eric Slack explains in an August 4, 2010 Storage Switzerland blog posting SSD in Legacy Storage Systems inevitably SSD performance runs up against the limitations of RAID controllers and storage architectures designed for much slower disk IOPs. Read the rest of this entry »
Long Live DRAM
DRAM JUST KEEPS GOING AND GOING AND…By Gareth Taube, Vice President Marketing, Kaminario
A recent Register article presents a good analysis of the pros and cons of SSD deployed as a storage tier, as a write or read cache, as a RAID array, or as a purpose-built appliance like the Kaminario K2. The article does a good job explaining that purpose-built appliances like the K2 are not constrained by the I/O limitations of disk and RAID architectures, but are instead architected from top to bottom with the throughput and advantages of SSD in mind. The result is superior performance. Read the rest of this entry »
Target the Business Problem, Not the Product
USE THE RIGHT TOOL FOR THE RIGHT JOBBy Gareth Taube, Vice President Marketing, Kaminario
It’s easy to forget sometimes that IT is as much about solving business problems and delivering business value as it is about technology. There are some interesting items in the blogosphere in the past few weeks about targeting specific business technology challenges when making IT purchase decisions.
In a May 18 Storage Soup blog posting, Randy Kerns cautions IT storage professionals not to get too caught up in the parade of storage events taking place over the next few months. It’s easy to get too enamored of the latest technologies after mingling with vendors and attendees, watching cool demos, learning about new product strategies and taking home all those white papers, case studies and other collateral, he says.
Instead, Kerns counsels IT to look inward first, focusing on the key storage related business challenges it will face in the next three to five years. Then, when you go to one of these events, have a specific set of requirements and evaluative criteria in mind for products and technologies that might address those specific challenges. He sums up his argument here:
There are two elementary starting points when looking at new products or technologies to meet the needs of an IT operation. You must ask if there is a capability that can only be done by using this product/technology, and if there is an economic advantage with a particular solution.