We have a DR FAS8020 (single node "cluster") system in another building that serves solely as a target for DR on our primary clusters (mirror & vault of a bunch of volumes). Over a hundred primary volumes each replicate out at some point during the day, and some subset of these would fail if the DR "cluster" was unavailable.
We're about to make our DR "cluster" unavailable from 2-6 hours in the near future to power the controller down and add an IOM12 module, and a DS212C shelf. While this system is powered down, aside from ugly messages in OnCommand/ActiveIQ, will anything else happen on the primary clusters that I need to worry about or plan for?
I'm guessing no, but since I'm not a storage guy, I figured the safest thing to do was ask.
This leads me to suspect that we don't actually need to bring down the single-node DR system after all - can anyone confirm that it is true that we can install an new IOM12 module into an empty IOM slot on a single node FAS8020 that has an active IOM6 controller and multiple IOM6 shelves?
Apologies, this has been a real comedy of errors. Our NetApp VAR spec'd the parts out for us and it looks like we've received the wrong stuff.
They told us we needed to add an "IOM12" module to the order to install into the FAS8020 since we only have IOM 6 ports on the controller (with 3 DS4246 shelves strung off the IOM 6), so we said "yea sure whatever" - so I was expecting that meant they were adding (what I'm guessing to be) a PCI-E 2x SAS 12GBps add-on card for the controller but nope, we got a X5720A in the mail which is for the disk shelf.
...So when we finally have the right part, I'm guessing we'll have to take a full outage if it is in fact a PCI-E card going into the slots shown here: https://photos.app.goo.gl/jC3UtnyAa2DgwXLt8. Not sure something going into those slots could ever be hot-swapped...