On Tue, Nov 28, 2017 at 4:59 AM, David Britton <[email protected]> wrote: > On Mon, Nov 27, 2017 at 04:22:40PM +0000, Jeffrey Lane wrote: >> This discussion makes me wonder, how many people do ISO based Server >> installs and would be affected by this? > > I think this is a good question.
Admittedly, a poorly worded on now that I read it again... what i meant to say was: "How many people do MANUAL ISO based server installs and would be affected?" In a previous life, I did a lot of installs to a lot of servers sometimes several times per day, and when I first started that team they DID, in fact, boot CDs and manually deploy each one. This was in a test lab that had several hundred servers, so you can imagine the amount to time wasted doing manual installs, even if you were doing assembly line installs across a full rack of systems at once (insert 30 CDs, Power on 30 servers, go to server 1, hit enter, server 2, hit enter ... server 30, hit enter, back to server 1, select Install, server 2, select install, and so forth). So my first update to the infrastructure was to set up a PXE environment to remove the need for booting via usb or CD. So the ISOs were then net booted. Then, installations were moved to kickstart or preseed and controlled by a simple shell script that would take care of pretty much everything automatically (it was a very ugly implementation of what MAAS looks like). Sorry for the confusion, I didn't mean to focus on the difference between ISOs and the MAAS images that are just blatted to the HDD, but more the difference between any sort of automated install vs a manual install where the user would actually hit this issue with having to catch a boot splash. In my own experience, that is just a very small, debugging corner case while the overwhelming majority of installs are done automatically, fully scripted and hands off to save as much time as possible. Anyway, as I said, it's more to satisfy my own curiosity about how others are actually deploying OSs in the wild. > Personally, when I have admined large datacenters with ubuntu, it was in > a test/dev environment, and use of the iso was helpful since the lab was > not controlled by MAAS, and the central PXE system was crude. > > So, this workflow was generic across anything that you could normally > boot from a CD, including things like FW update discs from HP/Dell: > > 1) Manual recovery/exploration of a single machine > > 2) If that worked, put that same CD image into the lab-wide PXE system > > 3) If that worked, look into the "cloud images" / MAAS approach on an > isolated vlan, if I could get one/justify it. > > -- > David Britton <[email protected]> -- Jeff Lane - Server Certification Lead, Tools Developer, Warrior Poet, Lover of Pie Ubuntu Ham: W4KDH Freenode IRC: bladernr or bladernr_ gpg: 1024D/3A14B2DD 8C88 B076 0DD7 B404 1417 C466 4ABD 3635 3A14 B2DD -- ubuntu-server mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-server More info: https://wiki.ubuntu.com/ServerTeam
