August 2010 - discuss - lists.mariadb.org

[Maria-discuss] Mac package (was Re: Contact Form - []: jeremy ellison)
by Hakan Kuecuekyilmaz 26 Aug '10

26 Aug '10

Hi Jeremy, all, On 16.08.2010, at 12:47, Bo Thorsen wrote: > Den 14-08-2010 05:25, Contact Form skrev: >> Looks interesting, why do you not have a mariadb package for Apple >> Mac OS X? This is the basis for my blog, bb, and server backend. If >> you have no Mac support it's not a drop in replacement for MYSQL. > > Hi Jeremy, > > The reason is that no one has offered to do it yet. MariaDB isn't a product sold by Monty Program, but an open source community project. Us Monty Program guys do a lot of the work, though. > > If someone will contribute the necessary work to get Mac packages, they will be included. Until then, it's source code only. Should you be interested in doing this, or know someone who will, then you should ask about the necessary work on the mailing list > maria-discuss(a)lists.launchpad.net. I read about your request for a MariaDB package for Mac OS X on my vacation and did some research. Back from vacation I have an alpha version of MariaDB package for Mac OS X for you and for our community to test. Caution: this is the first installer I wrote on Mac, so do not use it on a production system. Please test the installer and provide us with feedback. Known issues in the MariaDB installer: * The Preferences Pane app for starting/stopping the server instance is missing * The installer for setting up MariaDB as a Startup Item is missing. Side node: while digging into the Mac installer I found two bugs in the MySQL Mac OS X installer. * http://bugs.mysql.com/bug.php?id=56279 Mac installer does not work as documented * http://bugs.mysql.com/bug.php?id=56280 Mac installer's postflight script does not work all the time You can grab the package from here: http://lisas.de/~hakan/file/mariadb-5.1.49-osx10.6-x86.pkg Best, Hakan -- Hakan Küçükyılmaz, QA/Benchmark Engineer, Stuttgart/Germany Monty Program Ab, http://askmonty.org/ Skype: hakank_ Phone: +49 171 1919839

1 0

[Maria-discuss] Clustering with Maria/Aria engine
by Joshua Partogi 26 Aug '10

26 Aug '10

Hi all, Does Maria/Aria engine supports clustering? Or do we still have to use NDB for this? Thanks for your help. Kind regards, Joshua. -- http://twitter.com/scrum8

3 4

[Maria-discuss] mariadb-mysql command-line problem on CentOS 5.5 i386
by Daniel Bartholomew 25 Aug '10

25 Aug '10

I saw a comment today from Eric Koh on the Knowledgebase article for the mysql Command-line Client: http://kb.askmonty.org/v/mysql-command-line-client --------------------------------------------------------------------- mariadb-mysql command-line problem on CentOS 5.5 i386 I tried these 2 RPMs MariaDB-client-5.1.42-68.el5.i386.rpm MariaDB-client-5.1.49-82.el5.i386.rpm I get this message when starting the mysql client mysql: symbol mysql_get_server_name, version libmysqlclient_16 not defined in file libmysqlclient.so.16 with link time reference running ldconfig/rebooting doesnt help. The RPMs for the server are fine though. I am thus using MariaDB-server and MariaDB-shared together with MySQL-client-community --------------------------------------------------------------------- Anyone know what might be going on here? Thanks. -- Daniel Bartholomew Monty Program - http://askmonty.org

4 6

[Maria-discuss] my_atomic_add64
by Oleg Tsarev 24 Aug '10

24 Aug '10

Hello, I need my_atomic_add64 in mysql. Can i simple add following macros, or i need more advanced tricks? tsarev@main:/storage/project/percona/rtd_2$ diff -Nur ../rtd/c/include/my_atomic.h c/include/my_atomic.h --- ../rtd/c/include/my_atomic.h 2010-07-09 16:35:11.000000000 +0400 +++ c/include/my_atomic.h 2010-08-17 18:57:07.648819066 +0400 @@ -96,25 +96,30 @@ make_atomic_cas( 8) make_atomic_cas(16) make_atomic_cas(32) +make_atomic_cas(64) make_atomic_cas(ptr) make_atomic_add( 8) make_atomic_add(16) make_atomic_add(32) +make_atomic_add(64) make_atomic_load( 8) make_atomic_load(16) make_atomic_load(32) +make_atomic_load(64) make_atomic_load(ptr) make_atomic_store( 8) make_atomic_store(16) make_atomic_store(32) +make_atomic_store(64) make_atomic_store(ptr) make_atomic_swap( 8) make_atomic_swap(16) make_atomic_swap(32) +make_atomic_swap(64) make_atomic_swap(ptr) #undef make_atomic_add --- Oleg Tsarev, Software Engineer, Percona Inc.

3 4

[Maria-discuss] Decimal operations
by Oleg Tsarev 23 Aug '10

23 Aug '10

Hello, I don't understand how right work with decimal type. As you talk, i change column types from CHAR(...) to DECIMAL in my patch: http://www.percona.com/docs/wiki/patches:response_time_distribution Now i have trouble with decimal precision/scale. Here is my definition of system table INFORMATION_SCHEMA.QUERY_RESPONSE_TIME: ST_FIELD_INFO query_response_time_fields_info[] = { {"TIME", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, MY_I_S_UNSIGNED, "time", SKIP_OPEN_TABLE }, {"COUNT", MY_INT32_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONG, 0, MY_I_S_UNSIGNED, "count", SKIP_OPEN_TABLE }, {"TOTAL", MY_INT64_NUM_DECIMAL_DIGITS, MYSQL_TYPE_LONGLONG, 0, MY_I_S_UNSIGNED, "total", SKIP_OPEN_TABLE }, {0, 0, MYSQL_TYPE_STRING, 0, 0, 0, SKIP_OPEN_TABLE } }; So, i want use DECIMAL(20,6) for columns time and total, and DECIMAL(14,0) for columns count. 1) Does above definiton correct, or should include some information about precision/scale? I collect time of query execution in: 1) Column total - in int64, microseconds count 2) Column count - in int64, just count Here is source code, for fill this table: for(int i= 0, count= m_utility.bound_count() + 1 /* with overflow */; count > i; ++i) { const uint64 time = m_utility.bound(i); const int64 count = m_time.count(i); const int64 all = m_time.all(i); my_decimal time_d; my_decimal count_d; my_decimal all_d; my_decimal million_d; my_decimal temp_d; ulonglong2decimal(MILLION,&million_d); ulonglong2decimal(time, &temp_d); my_decimal_div(E_DEC_OK,&time_d,&temp_d,&million_d,SCALE); longlong2decimal(count,&count_d); longlong2decimal(all, &temp_d); my_decimal_div(E_DEC_OK,&all_d,&temp_d,&million_d,SCALE); fields[0]->store_decimal(&time_d); fields[1]->store_decimal(&count_d); fields[2]->store_decimal(&all_d); if (schema_table_store_record(thd, table)) { DBUG_RETURN(1); } } I want store some value to decimal type, than divide by MILLION: microsecond_count => convert to decimal => / 1000000 = result. But as result i get: SELECT * FROM INFORMATION_SCHEMA.QUERY_RESPONSE_TIME; TIME COUNT TOTAL 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 1 11 4 10 8 13 100 0 0 1000 0 0 10000 0 0 ...................... Instead of wanted: SELECT * FROM INFORMATION_SCHEMA.QUERY_RESPONSE_TIME; TIME COUNT TOTAL 0.000001 0 0 0.00001 0 0 0.0001 1 0 0.001 0 0 0.01 1 0 0.1 0 0 1 11 4 10 8 13 100 0 0 1000 0 0 10000 0 0 ...................... 2) What i do incorrect? How fix it? Best regards, Oleg

2 1

Re: [Maria-discuss] Full Replacement of MySQL (innoDB, mysql Workbench, data, etc)
by Daniel Bartholomew 18 Aug '10

18 Aug '10

On Mon, 16 Aug 2010 16:25:00 -0000 Enrique Daniel <enridp(a)yahoo.com.ar> wrote: Enrique> Hi !! Hello! I'm CC'ing this message to the maria-discuss mailing list in case anyone else wants to comment. I recommend you join the list. It's a good resource. Enrique> I'm really new with MariaDB, I have discovered it yesterday at Enrique> really :S And I have not tested it yet, but it looks really Enrique> promising. Thanks! We've done lots of testing to make sure MariaDB works as seamlessly as possible for those moving from MySQL to MariaDB. As with any upgrade, I would definitely recommend making a backup and testing MariaDB on your end to ensure everything works with your applications and use-cases before putting it into production. Enrique> Anyway, there are a few things that are braking me to make the Enrique> change. I hope somebody can help me: Enrique> Enrique> 1) I'm using mySQL Workbench to design my DB, how can we use Enrique> it with MariaDB? (Aria, XtraDB engine, etc.) I haven't used MySQL Workbench much, but using it with MariaDB should be exactly the same as using it with MySQL. Enrique> 2) My tables ar InnoDB, but I was reading about XtraDB and it Enrique> looks much better, how can I make the change from innoDB to Enrique> XtraDB? if I have a SQL file, is enough changing ENGINE=INNODB Enrique> with ENGINE=XTRADB (or something like that). Enrique> I want to preserve my FK, with cascades, and comments. For compatibility purposes, XtraDB identifies itself to the server as InnoDB, so there's no need to change ENGINE= in any SQL files. Any existing InnoDB tables will start using XtraDB automatically after the upgrade. Enrique> 3) What happens if I have my DB with data and working with Enrique> MySQL and InnoDB right now? can I change to MariaDB (and Enrique> XtraDB) preserving all my scheme, tables, data, relations, Enrique> comments, indexes, etc? Yes. All of our testing indicates you can upgrade from MySQL to MariaDB and all of your schemas, tables, data, relations, comments, indexes, and so on will "just work" like they did with MySQL without you having to change anything. Of course, as with any upgrade, I would recommend testing on a non-production server before putting MariaDB into production. One last note. If you do find anything that doesn't work, please file a bug report: https://bugs.launchpad.net/maria/+filebug Thanks. -- Daniel Bartholomew Monty Program - http://askmonty.org

4 5

Re: [Maria-discuss] Full Replacement of MySQL (innoDB, mysql Workbench, data, etc)
by Henrik Ingo 17 Aug '10

17 Aug '10

I'm forwarding this to the mailing list since of all the places it was sent to, lp did not send it to the list... Please take care of this, nice MariaDB people: henrik On Mon, Aug 16, 2010 at 7:25 PM, Enrique Daniel <enridp(a)yahoo.com.ar> wrote: > Hi !! > I'm really new with MariaDB, I have discovered it yesterday at really :S > And I have not tested it yet, but it looks really promising. > > Anyway, there are a few things that are braking me to make the change. > I hope somebody can help me: > > 1) I'm using mySQL Workbench to design my DB, how can we use it with > MariaDB? (Aria, XtraDB engine, etc.) > > 2) My tables ar InnoDB, but I was reading about XtraDB and it looks much > better, how can I make the change from innoDB to XtraDB? if I have a SQL > file, is enough changing ENGINE=INNODB with ENGINE=XTRADB (or something > like that). > I want to preserve my FK, with cascades, and comments. > > 3) What happens if I have my DB with data and working with MySQL and > InnoDB right now? can I change to MariaDB (and XtraDB) preserving all my > scheme, tables, data, relations, comments, indexes, etc? > > Thanks !!! > Enrique. > -- > This message was sent from Launchpad by the user > Enrique Daniel (https://launchpad.net/~enridp) > using the "Contact this team" link on the Maria Discuss > team page to each member directly. > For more information see > https://help.launchpad.net/YourAccount/ContactingPeople > -- email: henrik.ingo(a)avoinelama.fi tel: +358-40-5697354 www: www.avoinelama.fi/~hingo book: www.openlife.cc

1 0

Re: [Maria-discuss] Locale settings and fractal number
by Oleg Tsarev 16 Aug '10

16 Aug '10

Hi, Sergey! I use string column, because i need fixed-point number (time column): 14_digit_to_second.6_digit_to_microsecond 123456.654321 as example. Big number, because i have not only "time" column and also "sum" column too. I didn't find any type what good for this, and i use string. Typically this columns use only by hand or in scripts, and don't used in any another queries. When i prepare output, i should print number and i don't know what the delimiter should i use. What the internal function should i call for get correct delimeter? Best regards, Oleg. 2010/8/16 Sergei Golubchik <serg(a)askmonty.org>: > Hi, Oleg! > > On Aug 15, Oleg Tsarev wrote: >> >> Hello all, >> >> I have feature >> http://www.percona.com/docs/wiki/patches:response_time_distribution >> Now i want respect locale settings, and use in "time" column >> locale-depended delimiter instead of current hardcoded ".". >> I investigate source code about that, and also bugzilla, >> documentation, but not found answer to my question - where this >> delimieter setup. >> >> What the "official" mysql policy? Use only "."? Or use native-locale? >> Or use some internal mysql function? >> Please, resolve my confuse. >> I want understand how this questions solved in mysql. > > As far as I understand, MySQL typically does not uses system locale > information for formatting numbers or dates. > > Anyway, you shouldn't care about locale - as you have two columns > of numbers, you need to have two numeric columns, say, DOUBLE and > BIGINT. > > Regards, > Sergei >

3 2

[Maria-discuss] Locale settings and fractal number
by Oleg Tsarev 16 Aug '10

16 Aug '10

Hello all, I have feature http://www.percona.com/docs/wiki/patches:response_time_distribution Now i want respect locale settings, and use in "time" column locale-depended delimiter instead of current hardcoded ".". I investigate source code about that, and also bugzilla, documentation, but not found answer to my question - where this delimieter setup. What the "official" mysql policy? Use only "."? Or use native-locale? Or use some internal mysql function? Please, resolve my confuse. I want understand how this questions solved in mysql. Thank you very much. --- Oleg Tsarev, Software Engineer, Percona Inc.

2 1

[Maria-discuss] Google docs now allows collaborative graphics and schemas
by Henrik Ingo 15 Aug '10

15 Aug '10

Hi I didn't try it yet, but Google docs just advertised the addition of a collaborative graphics app and schema drawing app. It's not open source of course, but it's better than nothing. Especially if they turn out to support ODF fileformats. henrik (This was a topic discussed at the devmeeting in Iceland.) -- email: henrik.ingo(a)avoinelama.fi tel: +358-40-5697354 www: www.avoinelama.fi/~hingo book: www.openlife.cc

1 0