I've been working on onsite deployments for https://www.keystash.io, which is a Linux SSH Key and User management system. It's been going for a while now and I am finally implementing onsite deployments as so many customers actually want to run this themselves. When we started, we really thought customers wouldn't want the hassle of another piece of infrastructure to manage, guess we were wrong :-)
Onsite deployment is a lot more difficult to make slick and easy. We've been thinking about the best way for our customers to deploy while reducing the load on our support team. So far, we are thinking about RPM's, Debs and Docker and trying to make this as close to a '5 step process' as possible.
I would love to hear people's thoughts on other mechanisms that make it easier for SRM's / DevOps to manage key platform infrastructure software.
Not directly actionable but I would say very important: In the first year there is a major shift in priorities from yourself and as a couple to this new being in the mix. For my wife and I the lack of sleep, work stress and very little time to ourselves meant our relationship suffered. I would say with both our children, around the 9 month mark was probably the most difficult, with us at each other's throats a lot. The lack of sleep accumulated and made us quite irrational and short tempered.
To add to this we uncovered that my wife's hormones were out of whack until 2.5 years after our 2nd child was born. In that time she dealt with depression and was generally the most unhappy I have ever seen her (which means that extended to me). Our youngest is now 6 and I can say things have been normal and wonderful for a couple years now. We have so much time available now for each other and of course for sleep.
I openly tell people about my experience because I have noticed so many friends and colleagues go through similar experiences. They all felt like they were the only ones and they felt like failures (as did we). When I think back it's quite normal to go through this when I consider what was happening in our world and the amazing changes that happened to my wife's body and mind.
This reminds me when we were working on a business push notification app, much like WhatsApp business but back in 2015. Apple required users to allow push notifications for the app when installing. If a user disabled push notifications then it effectively stopped the app from working as it's sole purpose was to send push notifications. So to try combat users turning off push notifications without realizing what they had done, we would prompt them to say that turning off push notifications would stop the app from working. We submitted the app to the app store and Apple rejected the new version saying that we were forcing users to enable push notifications.
I remember being so frustrated with this process of trying to convince someone in Apple who just didn't seem to understand why this would make sense. They cited that it was a poor user experience but I can't imagine a worse experience than a messaging app that never received push notifications.
Onsite deployment is a lot more difficult to make slick and easy. We've been thinking about the best way for our customers to deploy while reducing the load on our support team. So far, we are thinking about RPM's, Debs and Docker and trying to make this as close to a '5 step process' as possible.
I would love to hear people's thoughts on other mechanisms that make it easier for SRM's / DevOps to manage key platform infrastructure software.