DDN (Datadirect Network) in production: what’s it actually like to live with day to day?
Been reading up on DDN (EXAScaler / Infinia / the AI400X appliances) and can't find much natural info, trying to separate the benchmark slides from reality. Hoping some folks who actually run it can chime in.
Mostly curious about the stuff you only learn after a year or two of owning something:
- Support. When something breaks at 2am, what happens? Do you reach someone who actually knows Lustre/the platform, or is there a lot of “please send logs and wait”? How’s the onsite field engineer side vs. remote? Any difference depending on region?
- Upgrades and fixes. Are code upgrades a non-event or a planned weekend with crossed fingers? How quickly do they turn around bugs once you’ve proven it’s on their end? Drive/controller swaps, smooth or painful?
- Overall verdict. Is it a solid product outside of the big HPC/AI showcase deployments? Is it complex to manage day to day without a dedicated Lustre person?
- Versus the others. If you’ve run or seriously evaluated VAST, Everpure, WEKA, or anything else alongside it, how did DDN stack up? Where does it clearly win, and where would you pick something else?
- Thinking of switching? If you’re on another platform and DDN is on your radar, what’s pushing you to look? Hitting a performance wall, support frustrations, pricing at renewal, or something else? And what’s holding you back from making the move?
Not looking for vendor or reseller input, please, and no DMs pitching alternatives. Just want honest takes from admin users and engineers, good or bad. Horror stories welcome too.
Thanks in advance.