HPE BladeSystem Review

Gives the administrator the ability to build a processing infrastructure remotely.

What is most valuable?

HPE BladeSystem c7000 is a complex piece of engineering.

  • I appreciate its simplified interface that gives the administrator the power to build a complete processing infrastructure remotely.
  • I like the on-board administrator as it gave me a very detailed set of information that allowed me to manage every aspect of the infrastructure remotely.
  • It allowed me to centralize the iLO remote access to every blade in the cabinet with very good performance.

You can view a Bladesystem like a modern car. The first thing you see is the body and the glossy paint, but under the hood, a bladesystem is essentially a group of servers (multi-core processors, RAM, Buses, storage, etc ), with redundant variable speed cooling blowers, redundant power suppliers, a large set of redundant connectivity options, and a big quantity of temperature and power consumption sensors, all of those connected and administered from a redundant administration module with many configuration parameters that you can accommodate in a bunch of ways to satisfy many different requirements.

Everyone of this modules are an appliance (a complete computer itself), and you can have a duplicate of the OA and the VC, just for redundancy and high availability purposes.

Before this brief description, probably you would agree that this is a complex architecture.

But the most attractive part of that , is that you deal with this complexity through a web portal that concentrate all of the configurations options, easy this tasks and guide the user with several wizards.

To manage all the parameters related to the Enclosure or Chassis, security access, and monitoring, you have to enter to the "onboard administrator module" (OA).

To manage all the aspect about LAN or SAN connectivity to the server blades you have to jump into the "virtual connect Module" (VC), but don't desperate, you have an hyperlink from the OA, that open the VC portal, to give you a seamless navigation between modules.

At last, but not least, you have the blades servers itself. You can have up to sixteen of those servers, with processors, memory, Out of Band management processor (Insight Lights-Out or iLO) and I/O cards (NICs, HBAs, CNAs, etc)

All of those have several Firmware (BIOS, NIC firmware, Power Regulator Firmware, HBA Firmware, iLO Firmware, Onboard Administrator Firmware, Virtual Connect Firmware, etc ) and you need to solve incompatibility issues between all of those.

The best part is that HPE give you an utility (HP Smart Update manager) that can manage all those firmware in a consolidated way.

HPE works hard to provide a centralized administration and good experience with the software, and if you are an advanced user, also can use an add-on to access all the configuration parameters using powershell (the administrator's task programming language that come in every windows operating system).

How has it helped my organization?

The first goal was to use blades. This stemmed from a space problem in our data center. We needed to add more servers, but the space became short quickly. The first consolidation approach was a blades server.

  • We administer all the systems remotely.
  • The blade server standardized more of our configuration processes with less manual intervention.
  • We also found that we needed less cabling. Now we can connect 16 servers in 10 rack units with 3 LAN and 2 SAN Fiber, instead of 48 LAN and 32 SAN fiber.

What needs improvement?

  • Hardware management could be improved.
  • Cisco UCS has a more universal approach. It treats the hardware as stateless and manages absolutely all configurations from the same console.
  • HPE has an on-board administrator to manage hardware aspects and virtual connect to manage LAN and SAN connectivity between blades and the rest of the world.
  • Firmware updates are complex. There are so many components and you need to account for the compatibility of all parts. Otherwise, you can have a blade that cannot start with the new firmware and then it takes extra time to solve the problem.
  • HPE has a utility for firmware updates that tries to provide peace of mind. It takes all these variables into account.

For how long have I used the solution?

We have been using this solution since 2008. We began mounting an 8 node VMware cluster. We began with one enclosure, a cabinet with 16 blade servers. We now have more than 18 of them distributed in different locations around the world.

What do I think about the stability of the solution?

We did not have any stability issues. The quality of the server itself enhances stability. Once the server is running, it runs for a long time.

What do I think about the scalability of the solution?

We don’t have issues with computer power scalability. We just add more blades, configure, install and go, or add more memory to an existing blade.

HPE also supports mixing several blade models in the same cabinet.

You can have, for instance, BL460c G7, BL460c Gen8, and BL460c Gen9 working smoothly in the same C-Class enclosure.

How is customer service and technical support?

In my country, the level of support is quite good. I recommend that you buy the server with a three or give year care pack to receive the manufacturer’s warranty.

Which solutions did we use previously?

I used rack form factor servers and switched to blades to gain consolidation ration. I also wanted to have better management control over the hardware infrastructure.

How was the initial setup?

The setup is easy if you install only one cabinet, and you know what to do and what to expect from the platform.

When you plan to grow your infrastructure to more than 16 blades, it becomes a little bit complex. You need to think about how to manage Virtual Connect Domains, MAC virtualization, and WWN virtualization.

If you design your platform based on that, everything will go fine. You will know what to do when a problem arises.

Which other solutions did I evaluate?

We have only used HPE infrastructure. Previous to blades, we were using Standalone Rack form factor servers like the DL380 model.

If I were brand agnostic, I would probably select CISCO UCS, but this didn't exist when we decided to use HPE blades.

Now, with Synergy Composable Systems, HPE probably takes a leap forward in technology and puts itself at the forefront. Please keep in mind that technology.

What other advice do I have?

  • The product is good and strong. Nowadays, the software works fine. I chose HPE because of the existing vendor relationship and reputation.
  • You can do all of the installation and configuration tasks. However, if you are not experienced, contract your first installation service to a partner with a lot of experience in that kind of equipment.
  • Take the corresponding training, as it is very useful.
  • I recommend taking the official HPE platform support. They have a BladeSystem course.
  • I also recommend users strengthen their knowledge with a Virtual Connect course. With this in mind, you can have a good experience with this kind of platform.
  • I also recommend that you progress further and think about automating the procedures, installations, and decommissions as much as possible.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Add a Comment

Sign Up with Email