Atlassian services are still down, 5 days and counting
old.reddit.com
old.reddit.com
I am honestly astonished that people traded on premise infra development for perceived stability of the cloud services. Not to mention all added complexity and lot of knowledge that will go down the drain as soon as you switch platform.
Unless you are there continuously to support such product, it can not last. Contrary to that, I have systems in production that nobody touched for half decade, having hundreds of thousands of users.
This is begging the question, though. I'll bet that the average company can't maintain the same uptime as a cloud solution, even with high-profile one-week outages like these.
> as soon as you switch platform
Implying that companies don't switch platforms when they have on-prem installations.
> I have systems in production that nobody touched for half decade
But that's apples and oranges. The comparison isn't Atlassian Cloud vs. your undefined production system, it's Atlassian Cloud vs. Atlassian On-prem (or $PLANNING_SOFTWARE_OF_CHOICE).
Its not only about uptime, its also about maintenance that is required to keep that uptime. It isn't there forever, you need to follow and adapt to API, license and other changes continuously.
> Implying that companies don't switch platforms when they have on-prem installations.
Be that as it may, you typically choose something that is not invented few years ago, but typically exists for decade and is battle tested. Lots of generic knowledge too, usable in other environments. And its not like people don't change Cloud platforms either...
> The comparison isn't Atlassian Cloud vs. your undefined production system, it's Atlassian Cloud vs. Atlassian On-prem (or $PLANNING_SOFTWARE_OF_CHOICE).
Comparison isn't between different supportive systems either. If I have a service, which is supported by specific infrastructure, I care about a service and current infra is just one mean to that end (I will happily replace parts of it that aren't compatible with that goal).
The comparison is a about which supportive system makes me do less work on the long run while keeping the same level of SLA. I don't care about individual features of supportive system in that regard, I care just about end effect. For one, if I have to relicense stuff every few months, its immediately not that good, as I can choose FOSS and let it be for decade.
I guess this is going to give them something to think about.
We have a similar story with our bitbucket, which is on-premise. And everyone wants it to move to cloud because it is falling behind in the feature.
I can imagine my jira or wiki beging down, but for the source code repository it can turn into a disaster very quickly.
Outline can replace a larger part of Confluence.