Kevin OBrien

Admin
  • Content Count

    1929
  • Joined

  • Last visited

  • Days Won

    38

Kevin OBrien last won the day on October 24 2018

Kevin OBrien had the most liked content!

Community Reputation

58 Excellent

About Kevin OBrien

  • Rank
    StorageReview Editor

Profile Information

  • Location
    Somewhere, USA
  • Interests
    Motorcycling, camping, geocaching, breaking electronics

Recent Profile Visitors

11640 profile views
  1. From what I can tell someone most likely breached an old account with an unchanged password, logged into the admin console here and sent out a bulk email. We've since pruned all the old admin accounts and reset passwords. It was sent through the bulk mail feature built into the forums itself, selecting all the users as recipients. So individual account as far as we can tell were not breached, just someone getting into an admin account. Right now the email engine on the server is disabled and the port is blocked at the firewall to prevent anything more from coming through while we check through all the layers.
  2. NAS options are going to be more feature filled, have standalone storage options rivaling more DAS offerings and allow connectivity into things like Glacier, etc. To match the speed of a DAS though you'd need to work in 10Gb which makes it a bit more expensive. If you don't exceed 100MB/s though you may not even notice a difference. So it comes down to how to use the storage, since it will be the defining element in which direction to go.
  3. Consumer SSDs still fall into the category of its great if they work, but when things start to become problematic with firmware, it can be usually pointed to use the appropriate drive for the application. Have you tried the light enterprise version of the 860 Pro? I wonder if the slowdowns are related the the over-provisioning differences between those two drives. Hit a write workload where the drive needs to run GC, and that could be the slowdown. To rule this out you would need to manually over-provision the SSDs with something like HDPARM, then put them back into the RAID group. Those S4500s
  4. I think the main part worth considering is the usecase. I like separating out direct attached storage and NAS, without as much intermingling the standards.
  5. Kevin OBrien

    Seagate BarraCuda SSD Review Discussion

    Thanks for bringing it to our attention, the SQL Server charts have the typo. I'll correct those soon.
  6. Kevin OBrien

    Samsung 860 QVO SSD Review Discussion

    There is no doubt they suffer under heavy or sustained write workloads, each manufacturer will tell you that from the start. They are really designed for micro workloads or read environments, which these tests are designed to show. For people looking at smaller workload footprints, the smaller benchmarks help more than showing a drive performing poorly in an area it was never designed to go into.
  7. Kevin OBrien

    Dell EMC VxRail P570F Review Discussion

    Its more towards the different type of test they are. VDbench will run full bore to leverage as much storage I/O as it can, while the SQL Server test we run is latency focused. You could easily drive more I/O, but its goal is more of a set I/O and aiming for lowest latency.
  8. Kevin OBrien

    Dell EMC VxRail P570F Review Discussion

    VDbench results surrounding database performance are geared more to show what full pedal to the metal performance potential is. Our SQL Server Benchmark Factory test is more useful at showing what a heavy load database does from a latency perspective. HCI traditionally will be a few steps slower than a SAN, but a SAN isn't the main competition. Its more towards integration, ease of use, etc while still being fast enough for most usecases.
  9. We we leveraging a 16Gb FC network (8 total ports out of the array). Switching was through two Brocade G620 switches
  10. Kevin OBrien

    Affordable & reliable RAID 1 controller

    Easiest might be... do you want a free RAID card from us?
  11. Kevin OBrien

    Eaton 5P1550GR UPS Review Discussion

    I added the runtime chart, Eaton lists out the variable timing based on load points.
  12. Kevin OBrien

    Intel Virtual RAID on CPU (VROC) Review

    We do track/collect CPU stats during the runs as part of the vdbench logging if you wanted to see more of the raw output. Its just not a part that we track.
  13. Kevin OBrien

    H700 vs 9265-8i

    If you have plans of only running hard drives, I don't think you will see any difference between those cards.
  14. Kevin OBrien

    Samsung 860DCT SSD Review Discussion

    The fishhook is the point in over-saturation where the drive decreases performance as latency increases.
  15. Kevin OBrien

    Intel SSD 660p Series Review Discussion

    Its not corruption, its a function of the test. Prior charts as posted are graphed on IOPS vs queue depth. The new charting method charts IOPS versus latency. With consumer drives as they get overloaded, latency spikes up and IOPS will actually pull back, causing the swinging motions. These are charted with VDbenches curve profile. unrestricted workload starts, measuring the "100%" performance of the drive, then it scales from 10-120%, plotting the bathtub profile up to and through peak saturation. Some drives peak more gracefully than others. Drive models that have very smooth and predictable lines are generally the highest performing. Drives that get bogged down and go nuts are the ones that lag on performance. For QLC drives we are actually working on an even smaller subset of tested drive surface to allow the drive to better cope with the workload. Those will be coming out soon.