| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comments:
<0> okay that means that it dpends on pma or something else
<1> lokus i see the error:
<1> 'Access denied. You need the REPLICATION SLAVE privilege for this operation'
<1> but I've done grant REPLICATION SLAVE on *.* to user@ip identified by "p***";
<2> Is it better to do one query with left joins if able to do so, rather than using several small querys? I mean, which one is faster?
<3> hy mysql is casee sensitive?
<3> .. case
<3> if i have a record "a"
<3> can i insert a value "A" in a unique set?
<1> lokus problem with the master, it hasn't the Repl_slave_priv column in user table of mysql DB :(
<4> splitt: what mysql server version
<5> alecs: If a value in a table is "A" and you do a SELECT * WHERE value='a', it will work. Its case insensitive for that
<5> alecs: not sure about inserting
<1> lokus 4.0.24
<1> the same in two severs
<3> i get the folowing result "Duplicate entry '0a' for key 1"
<1> but one has this columns(the slave) and the other one, no :?
<4> splitt: you upgraded from older version?
<1> lokus perhaps...
<4> splitt: run the mysql_fix_privilege_tables
<1> it doesn't affect to other tables, no?
<4> no
<1> i've to write an email... :)
<5> alecs: if you have an enumerated field like ENUM( "a", "b" ). I believe thats case insensitive too. So if you insert "A" it will be later be retrieved as "a". http://dev.mysql.com/doc/refman/5.0/en/enum.html
<1> that'll be all the problem
<1> thx a lot :)
<6> alecs: for 4.0, a/A are the same unless you are using a binary column or binary comparison
<5> Hey, whats the most efficient way of determining if there are any records in a table that match certain select conditions? Do I have to do a SELECT statement and then try to fetch a row and see if a row exists?
<6> Feldon: How else would you do it?
<5> Therion: wasn't sure if there was some all-in-one function for it
<6> You have to open the box to find out if the cat's dead
<5> Therion: I don't care about results, I just want to know if results exist
<7> Feldon: yes, if your select returns nothing then it failed, otherwise it'll return something
<6> SELECT COUNT(*) FROM table WHERE .........
<5> yah, ok :)
<6> Or you can SELECT 1 FROM stuff WHERE blah LIMIT 1;
<6> Which may get optimized rather more than any of the other options
<5> thx
<8> Question about database layout here, if that's okay...
<8> I
<8> er
<9> morning all
<3> re
<9> with DBD::mysql, will a date like "2006-2-1 00:00:00" work correctly, or do i need to add the leading zero?
<10> hey, is there an equivalent to : 'delete from node n where n.uid not IN (select u.uid from users u) ' for mysql 3?
<2> Is it better to do one query with left joins if able to do so, rather than using several small querys? I mean, which one is faster?
<11> phalcos depends on index and stuff
<11> inner join are faster then left join
<11> but using a huge inner join vs a couple of small queries you need to bench which is better
<2> So theres no "generally" in this case?
<11> usually its better with one join
<11> but that depends on how many rows are involved in the join etc
<11> and which index you have
<2> Say 10?
<2> all with the same index.
<11> usually only one index per table can be used
<11> so for a join the join index will be used (the index which matches the join)
<11> while for a small query then the index for the WHERE clause will be used
<2> Oh I see.
<2> I'll keep using small querys then.
<2> Thank you. :)
<11> mysql5 however have some mumbojumbo of constructing its own dynamic index behind the scenes but that seems to be a bit buggy currently
<11> or rather that the dynamic mumbojumbo doesnt boost the query as it should always
<2> Oh.
<2> Hehe..
<11> so its a performance "bug" not a true bug :P
<2> Okay.
<11> so best is if you create a couple of queries and run them to see which is best
<11> use the SQL_NO_CACHE to avoid hitting the query cache to get as good "real" results as possible
<2> Will do.
<2> :)
<11> like SELECT SQL_NO_CACHE blabla INNER JOIN 10 times blblaa vs SELECT SQL_NO_CACHE small query times 10
<11> also check the output of EXPLAIN SELECT from each method
<2> Ehm, you've given me alot more than I expected. :)
<2> So uhm, thanks. :)
<11> np
<11> but usually joins are faster than several queries
<11> but depending on the where clauses the several small queries method might be faster
<12> hmmm, the SELECT SQL_NO_CACHE thing is good
<12> I've been sending: ' SET SESSION query_cache_type = OFF ' at the beginning of a session
<12> and it doesn't appear to do anything
<13> ld.so.1: /usr/local/mysql/bin/mysql: fatal: libstdc++.so.5: open failed: No such file or directory
<13> echo $LD_LIBRARY_PATH = /usr/local/lib:/usr/lib
<13> what can be the problem?
<10> can I do a delete with 'left join' in mysql 3?
<14> ok, there's this problem I have which disappears with a sleep() between transactions
<14> transaction #1 updates a column
<14> commits data
<14> transaction #2 is opened in the sam process
<14> and checks for the data, before inserting some data
<14> the second check cannot 'see' the data inserted in the first transaction
<11> you need to change the transactionlevel
<11> tx_isolation or whatever its called
<11> http://dev.mysql.com/doc/refman/5.0/en/innodb-transaction-isolation.html
<14> Apachez: awesome, thanks
<15> i have a table mos_categories with id, title and section fields. another table mos_alias with url and target fields. i want to match on mos_categories.section values and grab all the mos_categories.id's which then will be used to find mos_alias.url rows using a '/content/view/$mos_categoriesSection/$mos_categoriesId.
<15> I am trying to have a list of rows of: mos_categories.title and mos_alias.url. is this possible with JOIN? (two queries or possible with one?)
<12> bah, if we have a really popular tag, like 20,000 records tagged, it's still going to be pretty slow to seach
<12> don't know if there's a way to optimize that... unless of course I stick it in a counting table like I'm doing with the "top tags for category"
<15> sort tags alphabetically to reduce cost in search
<15> that will give you O(n+k)
<16> whee
<11> isnt index sorted automatically ASC ?
<17> SELECT * FROM session_data WHERE sess_end < UNIX_TIMESTAMP()-3600 << This would return all rows that's older than 1 hour right? (sess_end contains a unix timestamp of when row was last updated)
<11> better if you use a constant
<11> like figure out what unixtime - 360 is in your script and use that value for the sql query
<7> not if there is a chance the app's time and the server's time are different
<17> I am not using any application/script for this... I am running the query in the command line tool...
<7> then that should work
<7> TIAS
<17> TIAS?
<7> try it and see
<17> yeah... but I wanted to be asure first... I will do a DELETE using that format on a table that contains 20,000 sessions... so I don't want to do anything wrong... many customers will be angry if so =) But now I know! Thanks for your help =]
<7> CopyNinja: that's why you do a select with that where first to make sure it's what you want
<7> then you replace the select * with delete when you're sure
<17> yup... that's what I did... but I also wanted to make sure with someone else before I made any conclusions wether it worked or not... because I don't actually know how many sessions that are outdated. but well... ill executed it and I hope it was right now =)
<12> csarven: ah, it's normalized so I'm getting the (indexed) tag name, and the resource_to_tag mapping table just has the tag_id
<18> good morning. I have mysql 5.0.18, innodb tables. Why "show table status" shows different number of rows all the time ? Tables are not touched by anyone.
<19> azzie, That's the way InnoDB is.
<20> count(*) is not maintained like myisam, its a guess
<18> will: thank you
<18> hmmm. If I do EXPLAIN SELECT ... ignore key (A) ... , why is it still showing A as Key ?
<21> http://home.flodhest.net/linux/mysql.txt?embed=true # anyone care to help me out with a SELECT statement?
<19> azzie, That it is using A, or that A is available as a key to use?
<18> will: possible keys: NULL, key: A
<22> is anyone here working with osx 10.4?
<23> Is it possible to combine two fields in a table like this
<23> SELECT * FROM table WHERE `firstname` `lastname` = Jerry Springer;
<24> k4st, yes
<7> Trixsey: CONCAT
<21> Trixsey, you need to use CONCAT(firstname,lastname);
<18> will: do you think it's normal or I should try opening a bug report?
<23> how do I get a space between first and lastname then?
<22> abre, did you use entropy.ch to get mysql? or did you use mysql.com?
<21> Trixsey, you need to use CONCAT(firstname," ",lastname); maybe?
<23> oh ok
<24> k4st, mysql.com
<22> okay, I will try getting it from there
<22> the problem is is that the entropy.ch thing has installed it
<22> but I don't know where it is
<22> and the spotlight won't find it either
<24> k4st, its in /usr/local/
<24> k4st, you'll have to use the terminal
<22> not very good with that, but I checked in camino by going to file:///usr/local and it's not there
<22> should it normally show there?
<24> its hidden
<24> you'll have tio use the terminal
<22> I've got it open
<22> typing in /usr/local/mysql in the terminal says the file doesn't exist
<7> try cd/usr/local/mysql
<7> err
Return to
#mysql
or
Go to some related
logs:
python urlopener.open timeout
kubuntu kscreensaverrc
#mysql
smtpd: bad command startup vhcs2
#perl
#openzaurus
ubuntu This codecs.conf is too old and incompatible with this MPlayer release!
#centos
Debian Do you have ESP Ghostscript installed?
#python
|
|