Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Databases

MariaDB 10 Released, Now With NoSQL Support 103

noahfecks (2379422) writes "Version 10 of the most famous fork of MySQL MariaDB has been released. Its developers said that is many times faster than MySQL, also claiming that its replications slaves are crash free. More details of this release can be found on the blog."
This discussion has been archived. No new comments can be posted.

MariaDB 10 Released, Now With NoSQL Support

Comments Filter:
  • "that is many time more faster than MySQ"

    That's the most awesome thing I've read all day. LOL. Or should I say, ROR.

  • by Anonymous Coward

    As if we needed any further proof that Slashdot has no editors.

  • by Anonymous Coward

    First all our bases, now this.

  • by AndrewFenn ( 1449703 ) on Monday March 31, 2014 @10:40AM (#46621827)

    I've tried out MariaDB specifically, the Galera Cluster many times and found it to be very lacking. The default Debian repos just seem broken and have been for a long time according to the bug reports i've read. Apart from the broken packages the fact that the documentation is very lacking and dotted all over the place has put me off. After MariaDB I moved on to Percona's implementation which comes with working packages and good documentation.

  • by Daniel Hoffmann ( 2902427 ) on Monday March 31, 2014 @11:08AM (#46622167)

    What does NoSQL mean for MariaDB? Without context it is just another buzz word.

    The only thing I know about MariaDB is that it is a fork of MySQL created because Oracle is Evil.

    • by Anonymous Coward

      NoSQL means the same thing it always means, "ACID is hard, so we don't do it."

      • by VGPowerlord ( 621254 ) on Monday March 31, 2014 @12:28PM (#46623019)

        NoSQL means the same thing it always means, "ACID is hard, so we don't do it."

        By that definition, MySQL is the original NoSQL database.

      • by lgw ( 121541 )

        NoSQL means the same thing it always means, "ACID is hard, so we don't do it."

        ACID is expensive to scale. NoSQL offers little when you have 1 or 10 DB servers. But if most of what you store doesn't need to be ACID, and you need 10,000 DB servers, NoSQL has a real cost advantage.

        • by Amadodd ( 620353 )
          Case well made for NoSQL - it's a great fit for 0.001% of use cases.
          • by lgw ( 121541 )

            As with everything cloudy, it's few software products but many, many end users.

            Well, that and the vast majority of simple programs that don't need ACID to begin with. If you just need a non-ACID structured data store, why bother with SQL? Currently NoSQL is mostly for analytics, but I think that's just habit.

            • If you just need a non-ACID structured data store, why bother with SQL? Currently NoSQL is mostly for analytics, but I think that's just habit.

              It gives you room to grow. If you're sure you won't need room to grow, that's fine, but it's still the typical answer to your question.

    • by Anonymous Coward

      The documentation server is going though the /. effect right now. But I was able to get a cached version off Google. Here is what I found:

      HANDLER
      The HANDLER statements gives you direct access to reading rows from the sto...

      HandlerSocket
      HandlerSocket gives you direct access to InnoDB/XtraDB and SPIDER.. It is i...
      Dynamic Column API

      Client-side API for reading and writing Dynamic Columns blobs

      Dynamic columns
      MariaDB starting with 5.3 Dynamic columns first appeared in MariaDB-5.3 Dyn...

      Dynamic Columns API
      This

      • The documentation server is going though the /. effect right now.

        Considering how weak the /. effect is these days it doesn't say much for the performance of their database.

    • by gbjbaanb ( 229885 ) on Monday March 31, 2014 @11:57AM (#46622759)

      no idea, but I know Postgresql has had JSON columns [postgresql.org] for a while now, so you get the benefit of 'typeless' data storage (ie a blob of JSON data) and all the benefit of relational data if you want it (as its just another column).

      MariaDB did it differently [mariadb.org], merging Cassandra as a storage back-end, and "dynamic columns' so you can have different columns of data per row in a table. (and you can get all the dynamic column data out as a JSON blob).

      • Hey that is interesting, I have a system here where I store small amounts of JSON data as text column, since I don't have to query that text column it is not a problem for me (although I have to deal with missing values by hand), but how would you go about querying data from those types of "document-store" columns?

  • Does MariaDB still do shit like this?

    http://sql-info.de/mysql/gotch... [sql-info.de]

    MySQL has always been a way to serve unimportant data at high speed. Great if you're serving up fuzzy matches to people who are doing a Google search and have no preconceptions about what they will get back in response to a search, or organizing a web forum visited by millions where if you lose someones comment, you really don't care. If you're dealing with data where accuracy, reliability and predictability is important, though, it was

    • by hholzgra ( 6914 )

      Depends on the SQL_MODE settings ... if using backwards compatible settings you'll at least get a truncation warning now, if using more strict modes it will throw errors instead ...

    • by neoform ( 551705 )

      I guess we better not use C/C++ in that case. Getting integer overflows means it's not a serious language, we should call them "fuzzy integers" instead.

      • I guess we better not use C/C++ in that case. Getting integer overflows means it's not a serious language, we should call them "fuzzy integers" instead.

        If you find a C++ compiler that will let you declare a variable of type 'foo' and store an object of type 'bar' in it without throwing any errors, despite 'bar' not inheriting from 'foo', then yes, I would say it's not a compiler you want to use to do serious work with.

        • by neoform ( 551705 )

          The first item from your link:

          He tries to insert 99999999999999 into a 32 bit int field, what he gets is 2147483647 stored as the value.

          What do you suppose would happen in C/C++ if you have a 32 bit int, and you add 99999999999999 to it? Are you going to curse C/C++ for allowing the int to overflow?

          Other databases (tested: Firebird 1.5rc4, Oracle 8.1.7 and PostgreSQL 7.4) raised errors with the same data.

          Allow me to introduce you to strict mode: http://dev.mysql.com/doc/refma... [mysql.com]

          Which has been available for

          • by ShieldW0lf ( 601553 ) on Monday March 31, 2014 @12:54PM (#46623325) Journal

            The first item from your link:

            He tries to insert 99999999999999 into a 32 bit int field, what he gets is 2147483647 stored as the value.

            What do you suppose would happen in C/C++ if you have a 32 bit int, and you add 99999999999999 to it? Are you going to curse C/C++ for allowing the int to overflow?

            Other databases (tested: Firebird 1.5rc4, Oracle 8.1.7 and PostgreSQL 7.4) raised errors with the same data.

            Allow me to introduce you to strict mode: http://dev.mysql.com/doc/refma... [mysql.com]

            Which has been available for almost a decade.

            You're like a dog with a bone. Last time I worked with MySQL was 5.0.1, and it was letting people insert ASCII strings into integer fields, and every time people expressed concerns, all you saw was rhetoric about how that should have been dealt with at the application layer. Which is fine if you're setting up a web forum, but not when you're organizing an enterprise that spans the world and has numerous applications accessing it, where one junior programmers mistake can hose your whole fucking enterprise. No client has mandated that I MUST use it since, therefore, I haven't used it since, and asked a serious question.

            The history of MySQL was very well summed up in an earlier post: "ACID is hard, therefore we don't do it."

            Not just me... most professionals know this and accept it and know that not every tool fits every scenario. Don't know what YOUR fucking problem is.

            • by neoform ( 551705 )

              >Which is fine if you're setting up a web forum,

              Or.. you know, a top 5 website [wikipedia.com]... small potatoes.

              >but not when you're organizing an enterprise that spans the world and has numerous applications accessing it

              Silly me, I didn't realize every tech industry could be summed up so tightly. Everyone that uses mysql must simply be retarded, and everyone that uses PGSQL a genius.

              >Not just me... most professionals know this and accept it and know that not every tool fits every scenario.

              And yet somehow you in

              • Re: (Score:2, Funny)

                by ShieldW0lf ( 601553 )

                >Which is fine if you're setting up a web forum,

                Or.. you know, a top 5 website [wikipedia.com]... small potatoes.

                Yes, exactly. A glorified web forum serving unreliable data at high speed. Do you have a reading comprehension problem or something?

                • by neoform ( 551705 )

                  Your definition of "web forum" is looser than mysql when it's not in strict mode.

                  • Try implementing hundreds of pages of ISO specifications for medical applications, then come back and talk to me about the "complexity" of Wikipedia. It has users, posts, edits, search and not a whole hell of a lot more than that. It's a web forum, not overly different from Slashdot. If they lose a post, no one really cares that much.

                    I suppose you think Slashdot is complex too, do you?

                    • by neoform ( 551705 )

                      Lol. I love how you're using a medical application as the standard/bar for all database use everywhere.

                      You would do very poorly working on large scale web services.

                    • Lol. I love how you're using a medical application as the standard/bar for all database use everywhere.

                      You would do very poorly working on large scale web services.

                      They seemed to like my work when I was helping rebuild all of Viacom's websites from scratch.

                    • I'm with neoform on this whole convo. You sound like an idiot from the first sentence. You used MySQL a long time ago and now you know every reason not to use it forever, right? Idiot.

                      Awh, man... and my self-esteem was all tied up in what you guys thought of me, too. Whoever the fuck you are.

      • C/C++ don't claim to follow relational data rules like MySQL does. Not only is SQL supposed to error if it can't do *exactly* as the user describes, it's supposed to change nothing if any of the affected rows error. It's not supposed to be allowed to guess if the user tells it to do something ambiguous or nonsensical. It's supposed to be required to throw an error in that case. Indeed, many RDBMSs error on some tasks simply because the result would be non-deterministic.

        An RDBMS is not just a fancy key-v

        • by neoform ( 551705 )

          >MySQL is the IE 6 of the database world. It encourages poor developer practice

          This would be true if IE6 was easier to use than every other browser out there.

          MySQL is simple, fast and does a sufficient job for those who use it.

          >It's not supposed to be allowed to guess if the user tells it to do something ambiguous or nonsensical.

          Considering most people who use MySQL also use PHP, this actually makes perfect sense. PHP is loosely typed, and in a way, so is MySQL. It's forgiving and makes corrections ba

          • MySQL is simple, fast and does a sufficient job for those who use it.

            I think his point is to say that MySQL is a full fledge SQL database is like saying IE 6 is a standards compliant web browser. Both do the job adequately for most people, but both aren't without serious faults.

            MySQL owes its success to web frameworks where better SQL servers like Postgres are considered an overkill and it works quite well in that domain. If your requirements are more on the SQL-side than the web-side of the equation, you

            • by neoform ( 551705 )

              >I think his point is to say that MySQL is a full fledge SQL database is like saying IE 6 is a standards compliant web browser.

              Herein lies the problem: what standards are you referring to? ACID? Mysql 5.5 in strict mode *is* ACID compliant.

              >MySQL owes its success to web frameworks where better SQL servers like Postgres are considered an overkill and it works quite well in that domain.

              I think you're remembering history differently than it was. MySQL was always significantly faster than Postgres, which

              • Herein lies the problem: what standards are you referring to? ACID? Mysql 5.5 in strict mode *is* ACID compliant.

                If that is really true then it's about time. It took them long enough. I knew something good was bound to come from Oracle.

                I think you're remembering history differently than it was. MySQL was always significantly faster than Postgres, which was a slow database until relatively recently.

                Nope. I remember history just fine. I remember having trouble with replication. I remember having to manually

    • by dtfinch ( 661405 ) *

      MySQL's had a strict mode since 2004 to reject invalid data. They didn't make it default until late 2012 though in 5.6.8, and I couldn't find what the MariaDB default is (short of downloading the source and looking). Even then, they only it in the default config file, so manual or distro-specific configs that omit the setting will fall back to the old truncation mode.

      • MySQL's had a strict mode since 2004 to reject invalid data. They didn't make it default until late 2012 though in 5.6.8, and I couldn't find what the MariaDB default is (short of downloading the source and looking). Even then, they only it in the default config file, so manual or distro-specific configs that omit the setting will fall back to the old truncation mode.

        Yeah, but it didn't always work. I know what the docs say. The last time I looked at them, they were wrong. I don't trust them not to still be wrong, because I've been in the trenches long enough know the man behind the project is a liar and an attention whore. I was hoping to hear from someone who could say "Yeah, I tested it recently, and the constraints work fine now/are still silently ignored."

  • Does it have schema-less mode?
    Have they buggered up the SQL parser?
    Have they buggered up relations?
    Can I separate out data sets and have 'em running on different node?

    • by mlk ( 18543 )

      It looks like it now has a schema-less (Key–value if I'm reading it right) mode as well as a method of dumping anything into said KV data store.

  • by MtHuurne ( 602934 ) on Monday March 31, 2014 @11:22AM (#46622357) Homepage

    The summary says the replication slaves are now crash free, but TFA says they are crash-safe. My database knowledge doesn't go very deep, but I think the latter means they won't lose data on crashes, not that they never crash.

    • No such thing as crash-free. Hardware/power/other software can cause crashes as well. Its way better to be crash-safe than crash proof. Crash proof is just waiting to be proved wrong.

  • This is new? (Score:4, Interesting)

    by Just Some Guy ( 3352 ) <kirk+slashdot@strauser.com> on Monday March 31, 2014 @01:19PM (#46623649) Homepage Journal

    I'd always thought MySQL was NoSQL to begin with. "Datatypes? Integrity? What geezer wants those! LOL! We're webscale!"

    (I love NoSQL DBs like Cassandra for the right applications. I haven't ever found an application for which I'd love MySQL.)

Whatever is not nailed down is mine. Whatever I can pry up is not nailed down. -- Collis P. Huntingdon, railroad tycoon

Working...