Harnessing the power of VMware vStorage APIs for Data Protection
Trucks move stuff. The engine burns fuel to generate power. Transmission system sends the power to the wheels. The trailer carrying the payload moves. Trucks consume a finite amount of time and energy to reach destination. We have learned in high school that engines and transmission systems have less than 100% efficiency. You are losing some of the generated power to heat and friction. Let us say you were given four of these big rig systems, all are identical with the exception of the transmission systems. Now you are measuring the time they take and fuel they burn to move a fixed amount of payload from one location to another. You repeat the experiments for different amounts of load. You plot the results to benchmark performance, efficiency and scale. Backup solutions integrating with VMware vStorage APIs for Data Protection (VADP) are similar to the transmission systems in trucks. The engine that really powers the backup process is the set of vStorage APIs. The goal of the backup solution is to harness its power to move data in virtual machines to secondary storage systems. Thus the virtual machines being backed up constitute the payload. As some of you already know, Veritas/Symantec had asked Principled Technologies to benchmark these transmission systems in harnessing the power of VADP. Here are the results. Remember that all trucks are equipped with the same engine (VADP). The load added (from 100 VMs to 1000VMs) for various tests are also the same. In accordance with the strict benchmarking standards, all the trucks were given a few runs to break-in so that the regular day-to-day operating conditions are evaluated. After the break-in runs, three benchmarking runs are conducted for each load. The median value for each truck for each loadis plotted above. Let us interpret the results. Performance The first thing that stands out from figure 1 is that the orange truck (NetBackup is represented in Orange. As this benchmarking started before the announcement of Veritas separation we used orange to represent NetBackup) is the fastest no matter how small or large the payloads are. And red truck (Competitor ‘C’) is the slowest! Even at the largest payload of 1000 VMs, NetBackup is 5.8x times faster than the slowest in the herd. The blue (Competitor ‘E’) and green (Competitor ‘V’) are weak when compared with the performance leader, NetBackup. However both are way better than the red truck. The bottom-line: If you are on blue or green trucks and run out of gas, do not hop onto to red truck! It is more like hopping on to an inefficient and slow steam wagon! Efficiency Now let us quickly gloss over figure 2. What you see is the CPU consumption at the data mover. In trucking terms, we are looking at how much gasoline is consumed for the trips and how much is still available in the reservoir if you had wanted to carry more loads. The orange truck (NetBackup) not only completed the trip much quicker but it also did the task with a lot of gasoline to spare. As you see can see we had to specifically draw attention to almost lost orange line in the resource utilization chart because of overwhelming gasoline consumption for pretty much all other trucks. The worst performer once again is the red truck. It is not just slow; it is also a gas-guzzler! There aren’t really any more resources to spare so you couldn’t have added more payload and hoped to finish the trip in the same time frame. The blue truck uses up nearly 85% of resources to for its trip. The greentruck starts somewhat well but the consumption shoots up during the final quarter. Scalability What exactly is scale in the context of this benchmark? The scalability is a measure of how well the product performs as the load it needs to handle increases. As you can see from figure 1, as you increase the load NetBackup continues to perform well ahead of the herd. But this data by itself only justifies that NetBackup is #1 in scale. Now let us move to figure 2. NetBackup data mover (media server) is using less than 50% of the gasoline while finishing the task in less than 1/5th of the time when compared to the worst performer. In other words, by the time the red truck moves 1000 VMs, NetBackup could have moved 10,000 VMs! That is 10x better scale! NetBackup is not only #1 in performance but it stays at that position while consuming the least resources thereby making more room for future demands. Thus NetBackup is indeed the king of scale. There is a lot more for us to share with you. How about scalability in recovery? How about storage integration? Will you be at VMworld San Francisco this year? Join George Winter and me at this session to learn more about the benchmark design, methodology, results and lessons learned. King of scale benchmark revealed in VMworld#STO54751.5KViews5likes2CommentsComparing apples to apples: All that glisters is not gold
Last week we made a shocking revelation. The storage snapshot integration from Competitor ‘C’ barely meets the RPOs that are already achievable in NetBackup Accelerator where operationally expensive storage integration is not required.546Views4likes0CommentsFree T-Shirts and Technical Stuff
As I've struggled my way through this last week on jet lag, limited sleep and caffeine at Symantec's Worldwide Sales and Marketing Conference, two things have struck me. The first is that Backup Exec has a special place in a lot of peoples' hearts. Its roots can be traced back around 30 years and there is a very real passion - among the sales force and pre-sales groups in Symantec, within the channel partner community and among our customers too. We brought themed "Backup Exec Boot Camp" t-shirts and hats to the conference as prizes and throughout the week more and more employees and partners were wearing them - through their own choice. Supporters if you like, but certainly a very loyal following that’s prepared to shout about it from the rooftops. If you want to get one for yourself, look out for the regional Boot Camps we're going to be running around EMEA in the coming weeks and months. Travelling around EMEA, I meet partners who have built their businesses around Backup Exec and customers who have been using it for years and who are genuinely pleased to meet "a Backup Exec guy" like me. The second thing that struck me this week was that in spite of the breadth and depth of capabilities within Backup Exec there is also a real awareness gap around those capabilities. I see focused competitors going after the VMware space - there's loads of stuff talking about them as a one trick ponies; there's more still trying to translate reality from marketing hype. I'm not going to dive into that right now, although Sean Regan wrote an enlightening bloghttp://bit.ly/kYxIQzthat I'd recommend you read. We’re doing our own bit when it comes to virtualisation, backing up more VMWare hosts than anyone else in the market, through Backup Exec and also NetBackup.(IDC market data & Symantec unit bookings). Here's the thing though. You could argue all day about the number of companies running Backup Exec, but it's a lot - well into seven figures. Many of those have already invested in the VMware Agent for Backup Exec in the three or so years that it's been available, while others are running the Hyper-V Agent in the same sort of time. Hyper-V platform uptake seems to be picking up more and more as well. Despite this I still run into customers and partners who don't realise these agents exist. There are others too. Backup Exec and De-Duplication, for instance. Symantec System Recovery - there's another example of how Backup Exec goes so far beyond tape-based physical server backup and can solve real problems for real organisations. My suspicion is that it's the approach we take in talking about it all. I don't just limit that to Symantec. It's endemic in the industry. We tend to talk about technical stuff - we all do it. What does it do? How does it work? My phone lets me check my mail and make calls (most of the time). I don't care how it does it; it solves my challenge of communicating with customers, colleagues, family and friends. So what problems does Backup Exec solve? It's simple really - it makes life easier. Whatever the question, the answer is Backup Exec. If my business is building boats, Backup Exec helps me build more boats. It lets me spend less time doing backup and restore stuff, less money investing in hardware, less time worrying about the complex bits and more time and energy building boats. Isn't that what it's all about? Or is it about free t-shirts and technical stuff? And just to clarify, Symantec is #1 for VMware backup, t-shirt or not.http://bit.ly/lXcWi6BackupExec 2012: New Info for the community.
There are a few new things worth checking out if you are in the Backup Exec Community: eWeek- Front page story about Backup Exec 2012 and VMware and Hyper-V. You have heard a lot about the new UI. This article and reviews explains a bit more about it and the focus on virtual machine backup and disk in the new release.