I am in the process of doing a controller swap from a FAS2620 to a FAS8200 using the Upgrading Controller Express Guide. The instructions say to netboot the new controllers to install the new Ontap version to match the version on the current controllers. I am netbooting to Ontap 9.7P5 and it has been over an hour and still loading the rootfs.img. It is proceeding as their are dots spewing across the screen slowly. I am direct connected to e0M on the controllers from my Mac at 1Gb speed. How long would I expect the netboot to take? Is there something I am missing?
e0M in FAS8200 is 1GbE. SP port is 100Mb/s indeed, but SP is not used for netboot.
I have seen this problem in the past when using http protocol; usually workaround was to use tftp. I even remember having seen KB regarding this issue, but now when NetApp destroyed their knowledge base it is no more possible to find anything.
Thanks. I will give tftp a try on our lab system. I was able to get the customer upgraded by just doing an Option 7 from the maint menu and then doing the swap. Had a couple of other challenges but able to work around them doing it that way.
@aborzenkov Which KB articles are you missing? We are always updating the KB site and publishing new info based upon cases and customer feedback, and if there is content you find missing, or not coming up in search we definitely want to know so we can fix it!
Looking further, it only seems to apply to FAS200/3000 series, not even 2xxx/32xx, so this isn't valid I believe. I apologize.
How far away is the HTTP server? I'm wondering if the performance is better or worse.
If you need help getting it working better, it may be faster to open a support case if you're having issues and we can help troubleshoot it. I'd personally want to get a packet trace to see if you are getting fast network response times.
I am direct connected from my laptop to e0M port. When I downloaded the software via Maint Menu - Option 7, it took less than a minute. So it appears to be a Netboot issue. I tried in my lab this week and looked like I had the same issue with http but tftp netbooted in a minute or two. So it definately looks like a netboot only issue with http.
My testing in the lab was with a Windows VM since I have not physically in the lab due to Covid stuff. The NetBoot from that VM was very slow using http and what I would expect from tftp. Once I can get in again, I will test both protocols using a direct connection from my Mac and try to collect packet traces for both. I will open a case at that point as well.