A Private Cloud with Vendor Lock-in Beats a Discombobulated Mess
The big news in the industry this past week was around the private cloud announcement made by and between EMC Corp., Cisco Systems Inc and VMware. In brief, these three companies are aligning to provide integrated virtualization product bundles for midsize, large and enterprise organizations that are referred to as Vblocks. Conceptually and practically, this is a smart move on the part of these three companies to deliver this type of service. Though some reports cited fear of user lock-in if this configuration is deployed, one would think concerns about deploying a private cloud that does not work as expected would be far greater.
Anyone reading through the pricing on these offerings is sure to be hit by an initial case of sticker shock. The low end Vblock 0 offering targeted at midsize enterprises for 300 - 800 virtual machines (VMs) starts at $100,000 while the upper end Vblock 2 offering aimed at enterprise corporations with 3,000+ VMs comes in at approximately $6 million according to an article that appears on searchstorage.com.
However I disagree with where the article goes at that point when it cites user fears about vendor lock-in. I have no doubt that they exist and I share those as well, but having worked as an end-user for over a decade and now working as an analyst and writer - guess what? The open systems model as it stands now where anything can theoretically connect to anything is irreparably broken, at least from a holistic perspective.
Companies need to take a hard look at how much value they are getting from their administrators when all they do all day long is troubleshoot configuration issues between Brand X's server hardware, Brand Y's network gear and Brand Z's storage system. The answer is not much.
A company may save some money up front by pitting HP and Dell against each other for server hardware or EMC and HDS for storage hardware. However the amount of money and opportunities that they squander managing these configurations and troubleshooting problems probably makes these $100K to $6 million price tags for the Vblock offerings pale in comparison.
Do not misunderstand, I am not suggesting that everyone run out and buy one of these Vblock configurations. Having worked with all of these vendors individually in the past, usually the only thing that works as well as they say it does at this point in the game is the printer that spits out the $6 million invoice. So when you start talking about these alliances and joint ventures between Cisco and EMC such as Acadia that 'accelerate customer build-outs of private cloud infrastructures', there is probably more that their support organizations and engineers don't know about to deploy private cloud computing in these organizations than they do know.
However something needs to be done to correct the open systems model of computing and I am convinced this is part of the solution. Is it perfect solution? Probably not. However is it a better approach than continually burning people out who are trying to make servers, network and storage work together in impossibly complex configurations? Absolutely.
So whichever HP, IBM, EMC-Cisco, or Symantec cloud computing architecture that a company buys into, it is likely going to be better than the discombulated mess they have been trying to nurse along for the last decade or so.
Besides, for those worried about lock-in, don't. The open systems genie is out of the bottle and none of these companies are going to put it back in. Like EMC's CEO Joe Tucci is quoted as saying in the searchstorage.com article, companies do not "have" to pick only their configuration. They can at any time go back to buying from multiple providers. But my sense is that many companies are willing to pay a premium for private cloud configurations that work reliably so they can get more value both the cloud and their administrators responsible for managing it.
This was not the only piece of news that caught my attention this week. I also had a conversation with a company called Storage Fusion regarding a change in the licensing of its Storage Resource Analysis (SRA) software. SRA in a nutshell provides companies a non-intrusive way to measure how much storage they have in their environment and then provides reports for it which is something companies should get their arms around before they do any sort of cloud deployment.
So what makes these guys different for ever other storage resource management products on the market? I noted three items. First, it charges a flat monthly fee regardless of how much storage you have in your environment. Overall I view this as better than the capacity, server-based and device-based licensing models I have encountered in the past, especially since they are not charging a fairly low amount for it.
Second, it uses no agents, collectors or any other devices. Rather, it provides scripts that users install on their existing servers that collect information from EMC, HDS, HP, IBM and/or NetApp storage systems, stores the data in a text file, zips them up in a text file and sends them to Storage Fusion for processing and analysis.
Third, it already is in use by consultancies like Glasshouse Technologies which gives Storage Fusion of added degree of credibility though it was unclear how pervasive SRA's use is among Glasshouse's clients.
Overall, Storage Fusion seems like a fairly low cost, low risk approach to gain some visibility into one's storage infrastructure. It's technique of deploying of scripts on individual servers that run on a regularly schedule makes for simple, low risk deployments and, while I did not have the time to thoroughly analyze this product, it seems to makes sense for companies to at least to take a look at this approach.
That's my weekly recap for this Friday! Have a good weekend everyone.
Anyone reading through the pricing on these offerings is sure to be hit by an initial case of sticker shock. The low end Vblock 0 offering targeted at midsize enterprises for 300 - 800 virtual machines (VMs) starts at $100,000 while the upper end Vblock 2 offering aimed at enterprise corporations with 3,000+ VMs comes in at approximately $6 million according to an article that appears on searchstorage.com.
However I disagree with where the article goes at that point when it cites user fears about vendor lock-in. I have no doubt that they exist and I share those as well, but having worked as an end-user for over a decade and now working as an analyst and writer - guess what? The open systems model as it stands now where anything can theoretically connect to anything is irreparably broken, at least from a holistic perspective.
Companies need to take a hard look at how much value they are getting from their administrators when all they do all day long is troubleshoot configuration issues between Brand X's server hardware, Brand Y's network gear and Brand Z's storage system. The answer is not much.
A company may save some money up front by pitting HP and Dell against each other for server hardware or EMC and HDS for storage hardware. However the amount of money and opportunities that they squander managing these configurations and troubleshooting problems probably makes these $100K to $6 million price tags for the Vblock offerings pale in comparison.
Do not misunderstand, I am not suggesting that everyone run out and buy one of these Vblock configurations. Having worked with all of these vendors individually in the past, usually the only thing that works as well as they say it does at this point in the game is the printer that spits out the $6 million invoice. So when you start talking about these alliances and joint ventures between Cisco and EMC such as Acadia that 'accelerate customer build-outs of private cloud infrastructures', there is probably more that their support organizations and engineers don't know about to deploy private cloud computing in these organizations than they do know.
However something needs to be done to correct the open systems model of computing and I am convinced this is part of the solution. Is it perfect solution? Probably not. However is it a better approach than continually burning people out who are trying to make servers, network and storage work together in impossibly complex configurations? Absolutely.
So whichever HP, IBM, EMC-Cisco, or Symantec cloud computing architecture that a company buys into, it is likely going to be better than the discombulated mess they have been trying to nurse along for the last decade or so.
Besides, for those worried about lock-in, don't. The open systems genie is out of the bottle and none of these companies are going to put it back in. Like EMC's CEO Joe Tucci is quoted as saying in the searchstorage.com article, companies do not "have" to pick only their configuration. They can at any time go back to buying from multiple providers. But my sense is that many companies are willing to pay a premium for private cloud configurations that work reliably so they can get more value both the cloud and their administrators responsible for managing it.
This was not the only piece of news that caught my attention this week. I also had a conversation with a company called Storage Fusion regarding a change in the licensing of its Storage Resource Analysis (SRA) software. SRA in a nutshell provides companies a non-intrusive way to measure how much storage they have in their environment and then provides reports for it which is something companies should get their arms around before they do any sort of cloud deployment.
So what makes these guys different for ever other storage resource management products on the market? I noted three items. First, it charges a flat monthly fee regardless of how much storage you have in your environment. Overall I view this as better than the capacity, server-based and device-based licensing models I have encountered in the past, especially since they are not charging a fairly low amount for it.
Second, it uses no agents, collectors or any other devices. Rather, it provides scripts that users install on their existing servers that collect information from EMC, HDS, HP, IBM and/or NetApp storage systems, stores the data in a text file, zips them up in a text file and sends them to Storage Fusion for processing and analysis.
Third, it already is in use by consultancies like Glasshouse Technologies which gives Storage Fusion of added degree of credibility though it was unclear how pervasive SRA's use is among Glasshouse's clients.
Overall, Storage Fusion seems like a fairly low cost, low risk approach to gain some visibility into one's storage infrastructure. It's technique of deploying of scripts on individual servers that run on a regularly schedule makes for simple, low risk deployments and, while I did not have the time to thoroughly analyze this product, it seems to makes sense for companies to at least to take a look at this approach.
That's my weekly recap for this Friday! Have a good weekend everyone.
Newsletter Signup
Leave a comment