Member Avatar for Mourin @ Dell

How are your current servers functioning?

Edited by happygeek because: moved to network forum
  • 7 Contributors
  • 8 Replies
  • 195 Views
  • 4 Weeks Discussion Span
  • Latest Post Latest Post by CimmerianX

Dani AI

Generated

This thread (March-April 2012) contains useful practitioner input from , , , , , , and . Their emphasis on reliability and service is foundational; the short commentary below highlights complementary operational, security, and lifecycle checks that are often missed at purchase time but determine real-world success over a server’s useful life.

Out-of-band management and automation: verify the vendor BMC feature set (remote console, virtual media, scripted power control, and an API such as Redfish or IPMI). During evaluation, use the remote console to run your installer and deployment scripts; confirm remote firmware update support, signed updates, and a tested rollback mechanism so updates can't leave systems unusable.

Compatibility and validation: insist on a short proof-of-concept with your actual workload. Validate OS and hypervisor driver compatibility, NIC offloads/SR-IOV or RDMA if needed, and storage paths (NVMe, NVMe-oF, SAS/SATA). Capture performance and power metrics under representative load so capacity planning, cooling, and density calculations are based on measured data, not vendor claims.

Security and lifecycle planning: prefer platforms with hardware root of trust (TPM, secure/measured boot) and a documented firmware/driver lifecycle. Agree on an explicit refresh cadence, spare-capacity strategy, and a firmware/driver support window with the vendor to avoid unexpected EOL problems.

  • Exercise remote console and virtual media; confirm scripted power cycles and remote re-image work.
  • Run a brief POC with representative workloads and record CPU/IOPS/latency and wattage.
  • Verify vendor driver/kernel support for your OS and hypervisor versions.
  • Test firmware update paths, rollback behavior, and whether updates are signed.
  • Confirm required network/storage features (SR-IOV, RDMA, NVMe) are supported and validated.
  • Define refresh cadence and a documented spare/repair strategy with the vendor.

Recommended Answers

All 8 Replies

Member Avatar for jbennet

parts availability
hotswappableness and modularity i.e. redundant psus, disks, fans ec...
build quality

Member Avatar for Coloradojaguar

Fast,low power use and high qualiity reliability from a trusted name and source.

Member Avatar for thelostboy

I would go with Parts, Hot swappable and tech support

Member Avatar for JorgeM

I concur with Coloradojaguar....reliability in the parts is key. You would also want to make sure that the vendor you choose uses the same parts in the specific model you purchase. Do not be surprised if you find that with some vendors, slightly different part are used in the same make/model. This makes it more challanging for you when you are looking for replacement parts. I would stick with a vendor, that has an excellent reputation, and good service options for you. From my experience this vendor list is very short.

Member Avatar for Mourin @ Dell

Thanks everyone for your feedback! Please let me know if you have any questions about the Dell PowerEdge servers.

Member Avatar for rubberman
  1. Easy access to the innards of the system, including disc drives.
  2. Good cooling.
  3. Fully buffered ECC RAM - this is error correcting RAM which will continue to run even if one of the memory sticks fail, allowing you time to take down the system cleanly before you replace the part(s).
  4. A big power supply (750-1000 watts).

I have an 8 core Intel server/workstation and these factors have saved my butt on more than one occasion. If you are going to run Linux, then make sure you install the lm_sensors module so you can actually monitor the CPU and RAM temperatures.

Member Avatar for Mourin @ Dell

Thank you all for your feedback! I've sent them to our planning and development team :)

Member Avatar for CimmerianX

1) 4 hour, onsite replacement service
2) Quality parts - especially drives
3) redundant power.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.