The title is far more dramatic than the actual content of today's post. This is an anecdotal story about a server I brought up and then took down, and I hope an insight into my personal planning process. Plus, thoughts about SharePoint.
A week or two ago, I got a rather harebrained idea to bring up an old physical server I have had laying around for a few months. The idea was to use it in part as a test bed for the next thing I want to do with my home network, and in part to use it as a low-watt heater.
I installed Windows Server 2012R2, patched it up, and then installed SharePoint and Office Web Apps 2010. I named the machine "Flying Kipper" and then joined it to the domain and gave it one of my public static IPs. Entered DNS information, and lo and behold, it was running. It was a great learning experience for the networking bit, actually.
It worked, but not particularly well. Ultimately it ended up working about as well as you'd expect SharePoint and web apps 2010 to work on a server from about 2003. It was just barely on this side of acceptable, and I think a few factors led to this.
The first is that the server is just old. It's got two 3.4GHz Xeon CPUs from what I believe might be the Irwindale generation. It is a very early implementation of HyperThreading. The SCSI disks are old and tired and probably not much faster than a modern spinning hard disk, and the 7 gigabytes of RAM may have limited the high end edition of SharePoint I was using, as well as the web apps.
Perhaps more importantly, the second factor is that the 2010 version of the web apps just work poorly in my experience. At the time of their release, they were more of a tech demo than an actual product. They are pretty badly coded in retrospect and they require SilverLight to work well, I had various problems using them in modern web browser, and it's a good look at the state of the art of web applications five years ago when they were new, but it's a bad productivity tool today.
I am willing to forgive old hardware its performance problems. My main server is already pushing five or six years old on its own at this point, and so there are a lot of newer and smaller services that can outperform it, or similarly sized boxes that can massively out-perform and out-capacity it.
Even ignoring the issue of security patches (which SharePoint 2010 is getting), I think there is a lot to say about using as modern a web application as you reasonably can, especially when more advanced functionality starts to be involved. In some browsers, the app didn't respond to keystrokes correctly, or it lost data in some situations where the backspace key was navigating to the previous page, instead of performing a backspace within the document.
To top it off, the whole thing doesn't support the multiplayer aspect to why I would even want to use it, as opposed to the current versions of Google Docs, Office Web Apps 2013/2016, Office 365, or the newest release of iWork, all of which support multiple active editors of a document.
I ended up, frustrated, turning off the server. The real main reason for this is that in a home the size of mine, there wasn't a way to escape the noise, ironically, except to go to the bedroom, which wasn't getting any of the benefit of having a server running.
Even though flyingkipper as stenoweb's new SharePoint server wasn't successful, it got me thinking about the next steps in the overall TECT migration, which is going to be to finish building up a new SharePoint server, with the web apps set up, and will be a big part of moving away from the classic all-integrated TECT.
In 2013, Microsoft split SharePoint and the newly renamed Office Web Apps into different roles. I'll be setting up maron and suddery as the SharePoint and Web App Server systems, respectively. One of the nicest aspects of the new architecture is that you can upgrade SharePoint and the Office Apps separately. That makes planning a little easier, and in an effort to get the most advanced functionality, I might even install the renamed-yet-again Office Online Server 2016 instead of OAC2013.
The next task is to come up with the structure for the sites on the new SharePoint server, and to get an SSL certificate for it. I'd like to look into Let's Encrypt for it, which should help the whole thing manage itself more easily. Maybe more importantly, I can't afford not just one but two or three new SSL certificates.
Although as of this writing, the blog hasn't made the jump to the new server, I'm starting to save documents to it, and eventually I'll actually start moving existing document libraries to it.
The real trouble will be, as always, keeping momentum up.