Quick warning. While rebooting an array for an update I noticed that it was configured to roll through 4 NIC’s worth of PXE on the boot order. If you have a PXE imaging server on the same subnet as one of the NIC’s this could cause the boot to hang, or worse if really misconfiguration re-image your array and destroy your data, or access to your data.
If your system is trying to PXE on reboot, please contact your vendor and discuss disabling it. While I noticed this on a Cisco Invicta, this isn’t really a Cisco problem and any modern software storage system this is possible. There is little more terrifying than watching a 300K storage device launch a windows PXE installer.
HDS has been on a long journey that has led us to “Year Z”. Almost four years ago the road map leaked to The Register that HDS would eventually merge their Modular (Then AMS, now HUS) with their Enterprise (VSP, now VSP G1000). The promise of a single block operating system, with a unified file, object, branch NAS and block management suite was a long way off.
Previously customers had to choose between platforms based on capacity, features, cost, up-time SLA’s, and performance. Often times a single feature requirement like storage virtualization would add a six figure amount to a bill of materials. Dependencies on High end ASIC’s made scaling down the costs of the VSP impossible. Today Hitachi has solved these problems and delivered a single platform that allows product selection to largely be done entirely based on capacity and performance needs. Features to seamlessly flow from the smallest to the largest platform on the line card. The G200, G400, G600, G800, G1000 provides a lot of sizes and price points without the confusion that multiple operating systems and system architectures provide. As other vendors add more platforms and OS variants to address different markets, its interesting seeing HDS consolidate product families. I’m curious if Netapp has some dusty old “One Platform” marketing slides that HDS can borrow.
My hats off to the engineers who managed to get full ASIC emulation running on the Intel Processors so that we can have VSP functionality without a six figure price tag. While I love the ugly duckling that is SNM2, it is good to see Hitachi moving on to faster, fancier management tools.
Infrastructure Director and the new management tools looks to match the “pretty” GUI’s that modern storage managers are coming to expect, as well as powerful automation and provisioning workflows to make provisioning and management a largely automated task.
I’ve loved the HUS for providing “simple” but reliable storage. I’ve often called it the pet rock of storage (configure, present everything to VMware, and stay in VMware for your management all day every day). VVOL support allows for snapshot offloading (Faster Veeam backups!) and more granular feature management. Most importantly it eliminates Vmware/Storage team miscommunication from causing VM’s to not get replicated, protected etc.