April 29th, 2011
Len Rosenthal We recently held our 1st Virtual Instruments regional customer council meeting in Dallas, right at the end of the Fall SNW conference. SNW was a big success for us as we had great traffic flow to our booth and our VI-sponsored customer case study presentation given by the Storage Team lead at JPS Health Network was attended by over 80 people. The customer council was a 1st for the company and we had participation from 6 existing customers along with a number of highly interested prospects.
Three of the customers (JPS Health Network, AT&T and Fidelity) discussed their use of VirtualWisdom in great detail, providing very interesting information on what problems they are solving and the key benefits of the VirtualWisdom solution. Thanks to the SNW venue, it was fortuitous that many of our prospects could hear directly from current VI users.
Overall, the feedback on VI was extremely positive. Comments like “We review VirtualWisdom reports every day. We couldn’t manage our SAN without it“, ”We saved over $1M by using the VirtualWisdom data to justify not buying additional cache for our storage arrays”, “We use VirtualWisdom data for capacity planning to optimize the number of servers that are able to connect to each storage port”, “We now TAP every storage port and strongly recommend that every enterprise data center with a FC SAN do the same”, and finally, “The data we get on CRC errors and aborts have been critical to avoiding and preventing outages. We found 20 bad SFPs within minutes of deployment. No other solution could give us this type of data”.
Of course, we have always known about these benefits, but it was great to hear them coming directly from our users. We also obtained some great input for future product planning.
April 22nd, 2011
Len Rosenthal With cloud computing – public or private – becoming more pervasive, companies who have made that leap are relying more than ever on their cloud providers to guarantee performance reliability and availability for their storage and storage networks, and to provide tools to measure those SLAs. Accurate, real-time visibility into the infrastructure can mean the difference between a complete system outage and a catastrophe averted.
As recent events have shown, when a major cloud provider goes down, it creates a cascade effect that impacts not only that service provider but all the companies that the cloud provider supports, resulting in service disruptions for multiple companies, negatively affecting business operations and customer perceptions. When this happens, businesses begin to question whether the benefits of moving their mission-critical applications to the cloud really outweigh the risks.
The problem with Amazon EC2 appears to be directly related to their SAN storage systems. The reality is the vast majority of problems related to storage and storage networking can be proactively identified and resolved before affecting users when the right I/O monitoring solutions are installed. What is the right solution? A true, real-time SAN I/O monitoring solution, one that is vendor agnostic, which can monitor across multiple domains with both fine- and coarse-grained metrics.
These virtual infrastructure optimization (VIO) solutions, such as Virtual Instruments’ VirtualWisdom® deliver detailed metrics that enable businesses to make smarter decisions when it comes to performance and utilization across every layer of their infrastructure from the network and storage to the application.
April 21st, 2011
Jim Bahn Would VirtualWisdom’s physical layer FC SAN monitoring have prevented the Amazon cloud outage? http://www.datacenterknowledge.com/archives/2011/04/21/major-amazon-outage-ripples-across-web/
Our customers often tell us that they’ve avoided serious problems by using VirtualWisdom to detect problems in the lower layers of the storage I/O stack. No one likes these kinds of stories written about them, so here’s a reminder to use our best practices guides to clean up your FC SAN. Check out this recorded webcast http://info.virtualinstruments.com/Webinar-Top-SAN-Errors-OnDemand.html, or this whitepaper on avoiding SAN performance problems: http://www.virtualinstruments.com/docs/WP_Avoiding-SAN-Performance-Problems.pdf. Better yet, schedule a call with a VI Solutions Consultant.
January 27th, 2010
Len Rosenthal Welcome to the VI blog on SAN best practices. This blog is our forum to share interesting new things we hear, tips and tricks, and just good ideas for improving fibre channel SAN performance , utilization and availability. Even though many of the folks here at VI have a unique skill-set around optimizing fibre channel SANs and virtual infrastructure, I predict that a lot of what we’ll post over the coming months/years will be things we learn from our customers. We’re constantly surprised at how they find new and creative ways to solve problems, and we hear from them that they’d like some way to share what they know, and learn from others.
I have to mention one thing to get us all off on the right foot though. The title “best practices” is more of a goal than reality. An old colleague once chastised me for calling something a “best practice.” He rightfully pointed out that what people were really looking for were “good” or “better” practices. Heck, most of us are just trying to avoid really bad mistakes. So in that spirit, I hope everyone feels free to comment and propose “other” practices, because if there was one best practice, we’d probably all be following it.
Fibre Channel SAN performance, troubleshooting, and optimization are considered somewhat of a “black art,” understood and successfully practiced by a chosen few. I mean, where can you go to get formal training? There’s no college degree program, it’s not even a sub-specialty in the computer science schools. Even the big storage vendors generally only have a handful of people who REALLY understand FC SAN performance. I’m talking about those folks who can read a fibre channel trace or for whom the title of “detective” is probably the best description of what they do.
The people I know who are really good at this share three attributes: (1) They’ve been working on this for years (2) They genuinely get a rush out of solving hard systems-level problems and (3) They’re really smart. So, while we have our share of these guys here at VI, I hope many of you will offer your experience and together we can increase our collective Wisdom about NetWisdom and VirtualWisdom.
Entries RSS