This is how I ended up doing it:
Initially I have been using IBOServices to do my local hard drive FDB
backups.
My routine that seems to work now:
1. Check for connected users using a query of MON$Attachments, other
than connection used in the utility app.
2. Disconnect the app I'm using.
3. GF
Hello:
After using Firebird (2.5 SS) exclusively in a desktop application
environment, I've created an associated app that stores data at an
internet address.
I just finished the code to allow me via a remote site to backup and
immediately restore as a maintenance operation the database.
My p
> what is the best way to do a join to a "top 1 row" select of second table? For
> example CAR and CAR_IMAGES to display just one image.
> I wrote following select, this works. However I think there must be a clever
> way to use "ROWS 1" instead of "min() and group by".
>
> SELECT
> CMRK.CAR_MAR
what is the best way to do a join to a "top 1 row" select of second
table? For example CAR and CAR_IMAGES to display just one image.
I wrote following select, this works. However I think there must be a
clever way to use "ROWS 1" instead of "min() and group by".
SELECT
CMRK.CAR_MARK_NAME,
CMDL.C
On Fri, Feb 3, 2012 at 10:25 AM, wrote:
>
> SQL databases are always in for a surprise about performance of simple
> statements.
Particularly true when you switch between record locking and MVCC.
>
> I have a table with the column 'ID' as BIGINT unique primary key.
>
> The table has about 4
thanks Ann,
You're right, the best option seems to be decimal (18,18).
(19,19) doesn't work :
SQL error code = -842. Precision must be from 1 to 18
--- In firebird-support@yahoogroups.com, Ann Harrison wrote:
>
> On Thu, Feb 2, 2012 at 6:50 AM, angel_ignacio_colmenares
> wrote:
> > I'm writing
Thanks for the suggestions.
For some strange reason all seems to be fine now. DB Workbench finds the
record
(and looks normal data) and works fine via ODBC as well. No idea what went
on there.
RBS
On Fri, Feb 3, 2012 at 6:43 PM, Thomas Steinmaurer wrote:
> **
>
>
> > Firebird 1.5 on Windows, acc
> SQL databases are always in for a surprise about performance of simple
> statements.
>
> I have a table with the column 'ID' as BIGINT unique primary key.
>
> The table has about 400 entries, the ID counts up without gaps.
>
> A simple
>
> select max(ID) from T
>
> takes about 14 second
> Firebird 1.5 on Windows, accessed via ODBC.
> Have a large table (some 50 million records) with the following structure:
>
> CREATE TABLE AUDIT_TRAIL
> (
>AUDIT_ID INTEGER NOT NULL,
>AUDIT_TIME DATE,
>USER_ID VARCHAR(12) C
> Hi, I'm trying to use fbtracemgr, but I get nothing coming out after I've run
> a query. Here is my config file:
>
>enabled true
> log_transactions true
> log_trigger_start true
> log_trigger_end true
> print_perf true
> prin_plan true
> log_procedure_start true
> log_procedure_finish
ruch...@gmx.de wrote:
>
> I have a table with the column 'ID' as BIGINT unique primary key.
> select max(ID) from T
> So Firebird seems to do plan a full table scan. (Hallo, anyone at home?
> I have an unique index on that column ;-) )
As far as I know, you need a DESC index for MAX queries.
Hallo,
SQL databases are always in for a surprise about performance of simple
statements.
I have a table with the column 'ID' as BIGINT unique primary key.
The table has about 400 entries, the ID counts up without gaps.
A simple
select max(ID) from T
takes about 14 seconds comple
Hi, I'm trying to use fbtracemgr, but I get nothing coming out after I've run a
query. Here is my config file:
enabled true
log_transactions true
log_trigger_start true
log_trigger_end true
print_perf true
prin_plan true
log_procedure_start true
log_procedure_finish true
and here is my c
On Thu, Feb 2, 2012 at 6:50 AM, angel_ignacio_colmenares
wrote:
> I'm writing a firebird provider for ServiceStack.OrmLite and unit tests
> ask for store the value : 0.686956521739130434782 in a table.
> the datatable is simple:
>
> CREATE TABLE TEST
> (
> v1 float,
> v2 float
> );
Float
> I'm trying with (to get all records from Feb 2nd)
>
> where
> (MyTimeStamp between '02-01-2012 00:00:00' and '02-01-2012 23:59:59')
>
> it uses the index, but I don't get all the records
My mistake!... that seems to work OK... Now I just wonder if that is the *best*
way to do that... So
hi,
using this to install firebird:
yum install firebird-classic.x86_64
it works in centos 6.2. now where is the ibase.h? i need that to compile some
other modules.
Thanks,
Wang
Hello! I have a timestamp field with an index
how can I select the fields from a given date, but using the index?
If I do:
where
((cast (MyTimeStamp as date)) between :from_date and :to_date)
I guess that is better not to do the cast and pass the params with date
+ time.
I use Delphi and
thanks Karol, I fixed it using 21 digits as you say.
now FirebirdDialect for Servicestack.OrmLite pass all unit tests.
btw, if someone is interested in using an ORM for Firebird can go to my
respository (fork):
https://github.com/aicl/ServiceStack.OrmLite/
or to the official one:
https://githu
Firebird 1.5 on Windows, accessed via ODBC.
Have a large table (some 50 million records) with the following structure:
CREATE TABLE AUDIT_TRAIL
(
AUDIT_ID INTEGER NOT NULL,
AUDIT_TIME DATE,
USER_ID VARCHAR(12) COLLATE NONE,
TAB
wang960 [2012-02-03 12:09] :
> Thanks for the reply, that answers most of the questions I asked. now, for
> the firebird client for OS X, doing it "by hand", any tips for me to explore
> this direction?
I think that you you need fbintl, Firebird, libib_util.dylib,
libicudata.dylib, libicui18n.dy
Thanks for the reply, that answers most of the questions I asked. now, for the
firebird client for OS X, doing it "by hand", any tips for me to explore this
direction?
--- In firebird-support@yahoogroups.com, philippe makowski wrote:
>
> wang960 [2012-02-03 02:53] :
> > Hi,
> >
> > Got three
21 matches
Mail list logo
