| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Comments:
<0> eIf: grant all PRIVILEGES on db.* to 'user'@'host' IDENTIFIED by 'somep***word';
<1> hi
<2> i have dialupadmin php interface with mysql-server , when i try to creat new user , i got error Could not connect to SQL database
<0> anyone here that could help me out :) ?
<3> !tell us about ask
<4> will[werk] asked me to tell you this: Don't ask if you can ask a question or ask if anyone knows about a particular topic. Just ask your specific question, then wait to see if someone will help.
<1> will!
<3> What?
<1> hi.
<0> lol
<5> What function will tell me how many bytes are in a BLOB?
<0> i have allready stated my question :>
<3> drewr, LENGTH?
<3> !man length
<4> (String Functions) : http://dev.mysql.com/doc/mysql/en/String_functions.html
<6> Is there a boolean data type in MySQL that I'm missing?
<6> or just use char(binary)??
<6> er, char(1) binary?
<3> minntc, Don't need BINARY.
<3> But that's it.
<6> k. how smart is MySQL's date data type? does it need a specific format, or will it read in different date formats correctly?
<3> It needs it in it's correct format.
<7> needs YYYY-MM-DD
<6> gr. okay.
<3> !m minntc date and time functions
<4> minntc: (Date and Time Functions) : http://dev.mysql.com/doc/mysql/en/Date_and_time_functions.html
<7> look at get_format though.
<5> will[werk]: Danke.
<7> and str_to_date
<6> buggy thing is I'm trying to gain access to this data in the raw from paradox tables. we'll see how that goes, I guess...
<6> for a currency type, use float(6,2)?
<3> No
<3> DECIMAL
<6> okay. oh yeah, fp math = teh ****
<3> I personally store money as cents.
<3> /100
<8> microunits!!!!!
<1> so query cash hit rate is 100%
<1> that's good, right
<1> i mean i have adequate cache specified
<9> dan__t, not only that but all your clients are selecting on material that is in the cache
<1> excellent
<9> are you using the mysql-administrator program?
<1> yea...
<9> one sec, there are some other good things to look at
<1> sure.
<1> I'm happy on the command line
<1> but the guys who pay me want to be able to see pretty graphs
<9> the graphs are nice, what is your key hit rate? 100%
<1> 0%
<1> heh
<1> "Average: 4"
<1> Sometimes I'll see something good.
<1> there's like a fsckton of keys in this database
<9> dan__t, that means your indexes need to be made better
<1> so much so, that I believe that they actually hinder preformance.
<9> for inserts?
<9> i have 70% cache 90% key
<1> I smell pastebin... brb
<10> hello guys
<1> squigly, http://pastebin.com/738389
<9> you really want to target your indexes
<9> i see there are no multi column indexes?
<1> I did not create these, I just started working here. On top of that, I don't know indexes too well.
<1> Correct.
<9> you want to get an example queury for each of your selects and then run them through describe and see how they work
<9> i cant see you needing all of them, there looks to be one for each table
<1> I've been using EXPLAIN on a lot of queries, yeah.
<9> *each coloum
<1> column ;)
<1> ok... let me toy with this
<9> like where you have select * from table where col1=value and col2 =value
<1> based on what you see, what kind of preformance hit would I be taking if I started to remove the indexes?
<9> an index of col1,col2 is best (mostely)
<1> Most of the carinality is quite high, so I'm thinking not that much.
<1> oh ok.
<9> it depends on what queries you are doing
<9> find test cases off all of your queries and have a go
<9> there is a slow query log
<1> it always reports nothing
<9> you can turn on and see what needs to be optimised first
<1> then again I need to increase the value...
<1> er decrease it, the timeout or whatever
<9> is it turned on properly?
<1> This server is just downright *bombarded* with queries
<1> Yes, but the timeout value or whatever it's calld, is bad
<1> or too hgih, not tuned
<9> lol, number of queries sent to server, 102326699 on my server
<1> i'm at 10.7M queries
<1> for about 6 days
<9> dan_t look at performance selects
<1> I'm sorry, I'll BRB.
<1> k sorry about that
<1> I think I can play around with some new indexes
<9> you really want to target them i think
<9> after look at your queries
<1> you're right.
<1> i'll do some reading on this.
<1> thank you.
<10> guys, I am having problems connection to mysql from asp: this is my reference line: strConnection = "Driver={MySQL};server=0.0.0.0;uid=root;pwd=p***wd;database=webhostpanel;" .... server, and p***word have the proper information. But the asp page is still complaining about a "missing datasource" and default controller not specified
<10> any idea?
<11> Darien you around?
<1> squigly, could a query potentially utilize more than one idnex?
<1> index, rather
<9> dan__t, yes
<1> select field1,field2,field3,field4 from table where field1 = "1" and field2 = "2" .... .... ....
<1> I could make an index on field1 and field2
<9> and it would be faster
<1> then field1 and field3
<1> then field1 and field4 etc etc?
<9> yes
<1> I'm starting to see the logic behind it
<9> but then inserts take longer
<1> that's why I don't want to have two columns in a multi-column index
<9> dan__t, its worth it
<9> but you dont want to have unused indexes
<1> I still don't know what an unreasonable value is for a multi-column index
<1> I would imagine that it would be highly data specific
<9> data and select specific
<9> in what you showed me, some thing like first_name,last_name might be a good one
<9> if you ever query on those together and directly
<1> what I'm having problems with is when the techs search for customers in this database
<1> they usually use lastname and zip code
<1> so it's only natural to make that a multi-column index
<9> lastname,zip
<1> but some use firstname, lastname, zip, phone number etc etc.
<9> yeah
<9> what does the code actually query?
<1> few bits of customer information - customer id, firstanem, lastname, zip, phone, city, state
<9> well you might want to make a mulit table query of all of those
<9> multcolumn index of all of those
<9> do you know what i mean?
<1> so CREATE INDEX index_name ON customers(firstname, lastname, zip ... ... ... ); ?
<9> when you run a seeach the code might look at first,last,zip,phone,city
<1> something of that nature
<1> yeah.
<9> run the actual query and show the SQL
<1> but that index on disk would be absolutely astronomical.
<9> how large is the table?
<9> i am suprised all of that data is in one table
<1> this is just the customers' table
<1> 40-something columns, 7 of which are used for the search query
<1> there's about 40 other tables themselves, with anywhere from 4-10 columns
<1> so it's laid out pretty well I think
<1> not normalized, but not terrible
<12> heya
<1> hi
<12> have a nice question for ya guys
<12> I noticed a big discrpency in file size when doing my monthly mysql backups today
<12> i found out it was because i have 2 servers and i was using mysqldump 10.10 for mysql 4.0.26
Return to
#mysql
or
Go to some related
logs:
#perl
installing pygame on suse
#web
#openzaurus
#dns
#css
sshfs motd
startxfce4 startx startkde
#gaim
suse can not open the file /proc/bus/usb
|
|