Technology Specialist at a energy/utilities company with 10,001+ employees
Vendor
2017-03-03T16:34:31Z
Mar 3, 2017
When I think about blades, I ask several questions, but start with the most important:
• Is important to think in your own technology roadmap first.
• The blades have two mayor components, the Chassis or enclosure, and the blade servers itself. Vendor Roadmap Longevity of the platform is the most important, to protect your investment. When you buy a chassis, you are setting up and infrastructure for at least 10 years or more. You need supportability from the vendor for this period of time.
• Centralized Management software for all of your blades Infrastructure
• Simplified operation and support in the future
Rack space, energy efficiency, noise, and different hardware blade configurations, are characteristics provided for all brands, in a similar way, so this is irrelevant to ask or consider these, when you establish a 10 years roadmap.
Infrastructure Architect at a retailer with 1,001-5,000 employees
Real User
2017-02-28T16:43:16Z
Feb 28, 2017
When looking at Blade Servers, the most important aspects in my opinion are resilience, scalability and support with a focus on TCO complimenting all.
You need to have a set roadmap when involving blade chassis and blade servers including your expected refreshes of the servers while leveraging the blade enclosure long term to lower Total cost of ownership.
This roadmap must contain the planned lifecycle of not only your servers, but be modular in approach for I/O interconnects to allow it to welcome and compliment future technologies.
Take the HPE C7000 as an example for medium sized enterprises, it supports half and full height blades for varying workloads. While, it has redundant network and fibre IO interconnects, it also includes additional interconnects to support future technology without requiring a forklift upgrade further stretching your budget by increasing ROI and reducing that TCO to the maximum.
As to support, you need to ensure you are getting the appropriate support level up front, if you have expectations for mission critical items, you cannot afford to get a low level of coverage. Ensure you get proactive care advanced, or datacenter care depending on the size of your datacenter footprint. This allows for proactive firmware/hardware assessments and provides you with your own TAM; a great value add to save the cost of an FTE where budget requires with larger deployments.
My opinion is my own and I do not have anything to gain from this statement above, I am an HPE customer currently, but have experience with other vendors as well.
Technical Sales Architect at a tech services company with 501-1,000 employees
Real User
2017-02-28T16:33:17Z
Feb 28, 2017
There is only one question I ask about blade servers - Will it make life easier as an administrator? Compared to pizza-box rack mount systems virtually all blades have similar physical capabilities, shared power, shared networking and shared management. But some systems are easier to deploy and manage. I focus on network simplification and comprehensive server management.
Traditional Blade architectures consolidated a block of rack systems to a single chassis - complete with switches for LAN, FC or IB - which adds complexity to my network and deployment tasks as I have to manage the cabling for each set of switches, track VLANs, VSANs, upstream connections etc. Network complexity grows with each added chassis. I want a system where the addition of a chassis is seamless, ideally I just add power and a few switch cables to a common fabric and don't have to worry about which switch is which. Like the model used for Cisco UCS.
I also want a server management system that completely abstracts the blade itself so its identity can be moved as workloads change, blades are added and so on, without having to reconfigure the network, tie old IP's to new UUID and other addresses, re-map PCIe blocks and so forth. The big problem with simply doing a Motion on my virtualized workload is that I cannot guarantee it will land on a new blade with the same VLAN's, VSAN's, security setup and so forth. My ideal server management tool has this kind of deep physical abstraction that allows me to manipulate my workloads without having to address these other issues. Again, like Cisco UCS.
Alliances & Channel Business Development Manager at Absolute Software
Vendor
2017-02-28T15:59:51Z
Feb 28, 2017
Consider the longevity of the platform. Remember the blade chassis is just an intelligent rack but it locks you into a vendors platform for at least your next 14 server purchases.
What is a blade server? A blade server is a minimal, utilitarian computer server that is designed in modular sections. The blade server is optimized to reduce the need for a larger footprint by reducing power consumption and removing certain components, while still having enough functionality to meet enterprise requirements. Multiple blade servers are housed in blade enclosures that deliver networking, power, management, interconnectivity and cooling. Instead of taking up horizontal rack...
When I think about blades, I ask several questions, but start with the most important:
• Is important to think in your own technology roadmap first.
• The blades have two mayor components, the Chassis or enclosure, and the blade servers itself. Vendor Roadmap Longevity of the platform is the most important, to protect your investment. When you buy a chassis, you are setting up and infrastructure for at least 10 years or more. You need supportability from the vendor for this period of time.
• Centralized Management software for all of your blades Infrastructure
• Simplified operation and support in the future
Rack space, energy efficiency, noise, and different hardware blade configurations, are characteristics provided for all brands, in a similar way, so this is irrelevant to ask or consider these, when you establish a 10 years roadmap.
When looking at Blade Servers, the most important aspects in my opinion are resilience, scalability and support with a focus on TCO complimenting all.
You need to have a set roadmap when involving blade chassis and blade servers including your expected refreshes of the servers while leveraging the blade enclosure long term to lower Total cost of ownership.
This roadmap must contain the planned lifecycle of not only your servers, but be modular in approach for I/O interconnects to allow it to welcome and compliment future technologies.
Take the HPE C7000 as an example for medium sized enterprises, it supports half and full height blades for varying workloads. While, it has redundant network and fibre IO interconnects, it also includes additional interconnects to support future technology without requiring a forklift upgrade further stretching your budget by increasing ROI and reducing that TCO to the maximum.
As to support, you need to ensure you are getting the appropriate support level up front, if you have expectations for mission critical items, you cannot afford to get a low level of coverage. Ensure you get proactive care advanced, or datacenter care depending on the size of your datacenter footprint. This allows for proactive firmware/hardware assessments and provides you with your own TAM; a great value add to save the cost of an FTE where budget requires with larger deployments.
My opinion is my own and I do not have anything to gain from this statement above, I am an HPE customer currently, but have experience with other vendors as well.
Aspects to look for while selecting a blade serer
1. Specific Application Requirements
a. Compute
i. Does the blade offer required cores ?
ii. Does the blade offer required flops
b. Storage
i. HDD, SAN or NAS capacity and connectivity
c. Networking
i. Required type & number of ports offered ?
ii. Required throughput offered ?
2. Infra requirements
a. Power consumption
b. Cooling requirements
c. Rack space & size ( of blade chassis )
3. Virtualization
a. Virtualization support offered for hardware
b. Required drivers etc.
4. OS requirements
a. OS platform supported on the blades
5. Other Hardware support
a. Specific NIC / HBA etc. requirements
6. Support
a. Ongoing support in case app being deployed in production env. with SLA etc.
7. Cost Comparison
Less footprint that is very important factor in the actual datacenters
Minimal Cabling
Easy to deploy and manage with OneView
There is only one question I ask about blade servers - Will it make life easier as an administrator? Compared to pizza-box rack mount systems virtually all blades have similar physical capabilities, shared power, shared networking and shared management. But some systems are easier to deploy and manage. I focus on network simplification and comprehensive server management.
Traditional Blade architectures consolidated a block of rack systems to a single chassis - complete with switches for LAN, FC or IB - which adds complexity to my network and deployment tasks as I have to manage the cabling for each set of switches, track VLANs, VSANs, upstream connections etc. Network complexity grows with each added chassis. I want a system where the addition of a chassis is seamless, ideally I just add power and a few switch cables to a common fabric and don't have to worry about which switch is which. Like the model used for Cisco UCS.
I also want a server management system that completely abstracts the blade itself so its identity can be moved as workloads change, blades are added and so on, without having to reconfigure the network, tie old IP's to new UUID and other addresses, re-map PCIe blocks and so forth. The big problem with simply doing a Motion on my virtualized workload is that I cannot guarantee it will land on a new blade with the same VLAN's, VSAN's, security setup and so forth. My ideal server management tool has this kind of deep physical abstraction that allows me to manipulate my workloads without having to address these other issues. Again, like Cisco UCS.
Consider the longevity of the platform. Remember the blade chassis is just an intelligent rack but it locks you into a vendors platform for at least your next 14 server purchases.
The most important features or aspect we was looking when we got blade servers are the following:
- Centralized management GUI that allow manage Network, profiles, servers, etc…
- Most efficient rack space use
- Most efficient energy consumption
- Blade systems allow to have different hardware configurations (CPU, RAM, HD, etc…)
Actually we use blade system from different manufacturers: HP, Huawei, Supermicro and Cisco.
Here we suggest go with HPE Synergy
Accelerate Application Delivery. Composable Infrastructure Platform.
Protect legacy IT · Turn ideas into value · World's first platform · Optimize applications
Lifecycle management efficiencies, including ease of provisioning, updating firmware, etc.