We are a small data consultancy, so we use other/more diverse things for customers, but our internal stack is fairly simple:
Hubspot/Jira/G Suite → (Python) → PostgreSQL → (SQL) → Tableau
Since we are Tableau partners, we have a bunch of internal licences either way. We host Tableau Server, ETL, and PostgreSQL ourselves, all on Centos cloud servers.
> It also doesn't look like a comfortable place for bikes. In fact, where are bikes even supposed to go on this street?
Another issue for cyclists is the permanent danger of getting your wheels in the rails, which means you are likely to either come to a sudden stop or, worse, come off your bike. There were experiments with "self-sealing" tram tracks that would lessen these risks, but no real solution exists quite yet, as far as I am aware.
It's a mistake everybody makes once, and then you learn to take tram rails at a sufficient angle. That's been my experience in Amsterdam. Also, when it's wet, there's the possibility of your rear wheel slipping on the tram rail if the angle is too small.
I'm not a big fan of putting my password (encrypted or not) somewhere where I don't have control. Therefore, I am using Passbolt[0] at work, since that gets me a browser addon plus web ui, while it also allows me to host it myself, i.e. where I can physically check what ends up written where in the database. Passbolt is open-source, encryption and sharing is GnuPG-based, and they have paid plans available.
Passbolt team was actually part of the Mailvelope project that did this. Integration setup is not easy / very user friendly, that's why it's not the default on Mailvelope.
"This guy" was Niklas Luhmann, by the way, an influential (and regularly misunderstood) German sociologist and social theorist, heavily inspired by Parsons and cybernetics.
Oh hey, maybe this will mean somebody at Tableau has to start to give a damn about enterprise features, such as a way to do product activation and registration that doesn't fail completely in non-persistent virtual desktop environments.
Or Tableau can continue to pretend that this isn't a real issue and stonewall customers and partners alike.
I think the actual visualisation part is neat, and better than many competitors, but many of the server-side parts are various levels of disastrous (as is their support), and their "data preparation" tool needs some serious improvements to be borderline useable.
15+ billion seems like a lot to me given how Tableau interacts with customers and partners alike, especially seeing how they are activelly alienating existing enterprise customers, all in favour of new sales, but perhaps something will change for the better here.
Yeah as the defacto Tableau Server administrator in my org, I am not happy with Tableau. Despite it's power and ability to let less-technical people generate insights from data, the administration experience as a whole is a total nightmare; purchasing/applying licensing to all installations is a disaster. Maintenance upgrades and backup procedures are also sort of wonky. For the data visualization piece, I personally feel like the software is difficult to use because it uses custom "branded" terminology for many things, so the knowledge you gain using Tableau can't often help you outside of Tableau... and then they place many features in locations that are not intuitive and thus difficult to find without documentation or prior experience.
We're a Microsoft-heavy shop, and I've been trying to get them to move to Power Bi simply because it's far more fully featured, easier to use if you're familiar with the "Windows way" of working, and has streamlined administration/installation/licensing/configuration in Windows environments.
I think Tableau Server works better on Linux, where upgrades etc. are also somewhat less painful and require less downtime.
That said, it baffles me why I have to restart Tableau Server 3 or 4 times during installation, and why I have to restart it for trivial changes more generally. For a piece of software that specifically ships with a cluster controller and full-blown zookeeper, somehow their engineers (or "engineers", as I sometimes get the impression) manage to make things that should be trivially solvable with reloads, partial restarts or spawning new workers (e.g. SSL certificates for the built-in Apache webserver) require a complete restart of the whole node.
edit: Regarding Power BI -- I feel that Tableau Server is (for better or worse) one of the killer features for many enterprise customers, because it means all of your data can remain within your own infrastructure and does not have to rely on external cloud providers. If that is not a requirement in your organisation, Power BI might make sense depending on your overall IT landscape, as well as your users' specific needs. On the other hand, if your organisation requires hosting things yourself, I guess it doesn't matter how miserable the experience is for you as an administrator. That's basically Tableau Server in a nutshell.
Note that a data lake does not necessarily replace the data warehouse, but rather often complements it. As such, you store your raw data from various sources in a centralised data store (Hadoop-like, NoSQL, etc.). From there, you prune, clean, select, and potentially aggregate data that you would like to provide in a quality-controlled way to your business users, in a data warehouse. This data warehouse most often will be a more traditional relational data store (usually some flavour of SQL database), which allows users to select data from a curated, pre-selected slice of the overall data stored in the data lake, and which enables easier integration with common reporting tools, whether more traditional standard reporting tools or self-service BI tools.
Kyle Rankin gave a (very good) keynote at FOSDEM 2019 about this topic, harkening back to the Unix wars, Sun etc. He has some really nice one-liners. Video is available here:
There is a "suggestion" somewhere that asked them to implement service accounts for API access. They essentially replied with "this is a good idea and we should do this". That post was from 2012 or so, and we still don't have service accounts.
openSUSE still "uses" YaST, as in, if you choose to change something via YaST, it will happily do so. You can also simply install/uninstall various components. It also happily integrates with changes made elsewhere (e.g. the "Firewall"-module wraps around firewalld nowadays, which you can influence either via YaST or firewall-cmd etc.).
I sometimes use the Package Management part of it, since searching for something is sometimes quite nice in it. That said, I don't use it for installing software or doing updates etc., so you can absolutely use e.g. zypper & YaST side-by-side.
Correct. I'd say openSUSE is more akin to Fedora in that context as both base and development distribution for SLES, much like Fedora is for RHEL (which in turn then "becomes" CentOS).
There are 2 OpenSUSE variants, LEAP and Tumbleweed. Their release model has changed maybe 3 years ago.
LEAP is exactly the same package versions as in the newest SLES I believe. I would expect them often to be a bit older package versions than Fedora, but I haven't compared recently.
Tumbleweed is a leading/bleeding edge rolling release distro. The closest equivalent is Fedora Rawhide.
Hubspot/Jira/G Suite → (Python) → PostgreSQL → (SQL) → Tableau
Since we are Tableau partners, we have a bunch of internal licences either way. We host Tableau Server, ETL, and PostgreSQL ourselves, all on Centos cloud servers.