Are you interested in Analytics? In Databases and Big Data? BI? AI?

[JeffKeryk]

Welcome to my world. Here's all ya need. By the way, this is how just about everything on the WWW, business, etc. works; it uses data. That data must be in a format suitable for the target application. There is no 1 size fits all.

𝗞𝗲𝘆-𝗩𝗮𝗹𝘂𝗲 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : Picture a colossal locker room. Each locker (value) has a unique key. Storing and retrieving data becomes a cakewalk! Examples: Amazon QLDB, Redis, AWS DynamoDB.

𝗦𝗽𝗮𝘁𝗶𝗮𝗹 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : Geography and technology entwine! These databases store geographic data (like coordinates for landmarks or cities). They're the powerhouse behind Google Maps and Uber!

𝗟𝗲𝗱𝗴𝗲𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : These databases stand for transparency and immutability. They're the backbone of blockchain technologies, with Apache HBase being a notable example.

𝗡𝗲𝘄𝗦𝗤𝗟 : It's the perfect blend of NoSQL scalability with SQL's reliability. Key players include CockroachDB, VoltDB, and NuoDB.

𝗦𝗤𝗟 (𝗥𝗗𝗕𝗠𝗦) : The classical choice! SQL databases store structured data in tables and support powerful queries. MySQL, Oracle, and Microsoft SQL Server are stalwarts in this space.

𝗚𝗿𝗮𝗽𝗵 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : The social network of databases! They're champions in delivering deep insights about connections. Think LinkedIn network analysis with Neo4j or AWS Neptune.

𝗖𝗼𝗹𝘂𝗺𝗻𝗮𝗿 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : These databases store data by columns, not rows. Ideal for analytical queries and efficient data compression. Apache Cassandra and DataStax are well-known examples.

𝗧𝗶𝗺𝗲-𝗦𝗲𝗿𝗶𝗲𝘀 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : A dream for tracking changes over time, like stock prices or weather data. InfluxDB and Prometheus are top picks for their superior query performance.

𝗗𝗼𝗰𝘂𝗺𝗲𝗻𝘁 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : Store data in a flexible, document-like format (similar to JSON). MongoDB and Couchbase excel in this space, offering efficient querying and flexibility.

𝗢𝗯𝗷𝗲𝗰𝘁 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲𝘀 : Data as objects, mirroring object-oriented programming. ObjectDB and ZODB provide encapsulation, inheritance, polymorphism, and more.

 

[Pew]

I'm glad my only interaction with SQL is installing it for other users

 

[Slick3914]

It all comes down to “what number do you want?”

 

[wwillson]

The BITOG backend is MySQL, which is a phenomenally reliable database. The SQL server answers, on average, about 350 queries/second. That's about 11 Billion queries annually. We've been on MySQL since about 2003 and I don't remember ever having a problem with the DB.

 

[cptbarkey]

My company played with graph for a few years but then we converted it to ms sql. Maybe our implementation wasnt correct, but its intended use (hierarchial information) didnt deliver, and support is expensive. Other then that, most of my career has been thanks to ms sql and oracle. Databases and data is fun!

 

[Skippy722]

All I know is I should have gone to school for computer programming instead of machining

 

[MolaKule]

My son is an SQL and DB2 database expert and does analytics all day.

 

[ripcord]

I don't do much analytics, but I spent all day yesterday optimizing a gi-normous SQL stored procedure and fighting with an unfortunately stupid DB design.

 

[alarmguy]

All I know is I should have gone to school for computer programming instead of machining

I have a family member, genius sort of. Started with programming, got bored, went to WallStreet became wealthy as a data analyzer or something, now reports directly to the CEO of one of the country's largest growth companies of the last decade. Not the largest but a name everyone in the country knows.

 

[JeffKeryk]

It all comes down to “what number do you want?”

In finance, we tell an old story...
3 CFO candidates come in for an interview.
Question: What's 1 + 1?
Candidate #1: "2".
"Thanks, we'll get back to you."
Candidate #2: "3".
"Thanks, we'll get back to you."
Candidate #3: "What do you want it to be?"
"You're hired!"

 

[JeffKeryk]

My son is an SQL and DB2 database expert and does analytics all day.

Very cool. Tell him I started with SQL 6.21a which is when Microsoft split from Sybase. As a Microsoft insider, I was invited to preview Power BI a year before release.

Ask him about Data Science; Google is starting Data Scientists right outta school for $300K plus stock. A great time to be in the field.

 

[JeffKeryk]

My company played with graph for a few years but then we converted it to ms sql. Maybe our implementation wasnt correct, but its intended use (hierarchial information) didnt deliver, and support is expensive. Other then that, most of my career has been thanks to ms sql and oracle. Databases and data is fun!

No implementation is perfect and every one is filthy expensive. My rule is allow for change; flexibility is key to long term success.
If the implementation is good, most of what the business thought they wanted quickly goes away; the next set of needs are far more difficult.

 

[JeffKeryk]

The BITOG backend is MySQL, which is a phenomenally reliable database. The SQL server answers, on average, about 350 queries/second. That's about 11 Billion queries annually. We've been on MySQL since about 2003 and I don't remember ever having a problem with the DB.

I believe that speaks to good design more than back end RDMS. Just my experience.
Database applications live and die on 2 measures: uptime and speed. Impressive!

 

[JeffKeryk]

All I know is I should have gone to school for computer programming instead of machining

It takes all kinds and all work is honorable. I'm sure you are great at what you do and bring home the groceries!

 

[andrew_j]

I love databases and been working with all sorts since 1990.

 

[wwillson]

Tell him I started with SQL 6.21a which is when Microsoft split from Sybase.

I used that version a whole lifetime ago.

Power BI

OMG do I love PBI. We have huge spreadsheets at work that will just stall when you try to do any operations on them (CSV exports from databases). I import them to PBI, then do joins to make a super useful report. PBI will query, sort, filter, join huge tables, where excel will just tip over. I've become somewhat well known inside the company for making reports that are actually useful. It's not really that hard folks, just pick up a new toy and play with it until you figure out how to make it do stuff for you. PBI does stuff really well.

 

[JeffKeryk]

OMG do I love PBI. We have huge spreadsheets at work that will just stall when you try to do any operations on them (CSV exports from databases). I import them to PBI, then do joins to make a super useful report. PBI will query, sort, filter, join huge tables, where excel will just tip over. I've become somewhat well known inside the company for making reports that are actually useful. It's not really that hard folks, just pick up a new toy and play with it until you figure out how to make it do stuff for you. PBI does stuff really well.

Perhaps you understand PBI uses the MS SQL engine under the hood? You are a dB guru!
Did you consider importing directly from the dB source? Please remember, Excel is evil. No security, you can Micky Mouse the numbers, stale data, etc... Of course security is only everything.

 

[wwillson]

Perhaps you understand PBI uses the MS SQL engine under the hood?

Yep!

Did you consider importing directly from the dB source?

That will never happen in a million years. Nobody outside of the dev and maint teams are allowed to even look at these databases. Huge beyond huge and most run on Daytona. If you know what Daytona is, then you are really cool

 

[supton]

I don't know anything about databases. All I know is that Excel makes a lousy database if you have to share it with others (problems with multi-access). At some point I need to learn something about SQL.. but all I need is one step up from Excel.

Analytics would be nice, but then I'd have to have data going out and back into the database... and for that, I want nothing to do with. Supposedly we're getting a PLM system and I'm hoping any analytics can be handled there, for me, by someone else.

 

[JeffKeryk]

Yep!


That will never happen in a million years. Nobody outside of the dev and maint teams are allowed to even look at these databases. Huge beyond huge and most run on Daytona. If you know what Daytona is, then you are really cool

Actually, I misspoke. PBI generates a multi-dimensional cube, which is SQL Server Analysis Services, aka SSAS. A thousand pardons.
Daytona is a high speed race track in Florida? A 68 L88 Vette, the Sunray DX roadster, hit over 180 MPH on the bank...