Hi, is this a correct behavior?
If I prepare statement with a limit statement, e.g. *SELECT * FROM
system_range(1,10) LIMIT ?* and then get the parameter meta data, I would
get VARCHAR as a data type of the limit related placeholder. I am expecting
a numeric data type, because *LIMIT 'abc' *is
update: the same behavior applies to OFFSET ?
Dne pondělí 9. března 2015 10:43:32 UTC+1 Vojtěch Hordějčuk napsal(a):
>
> Hi, is this a correct behavior?
>
> If I prepare statement with a limit statement, e.g. *SELECT * FROM
> system_range(1,10) LIMIT ?* and then get the parameter meta data, I wou
Ah, I see :S
Sorry, SQL blindness. I think especially when it comes to DDL I will copy
something and once it works not worry about the SQL (I mean it is unlikely
I put the NOCHECK in on purpose).
Thanks for taking a look.
On Thu, Mar 5, 2015 at 5:45 PM, Noel Grandin wrote:
> On Thu, Mar 5, 20
Hi,
I have a query performance problem that has come, gone away and come back
again.
The issue seems to be when using what are sometimes referred to as sub
types. A sub type is a table that share a primary key with the super type
table. This explains the scheme a bit better.
http://blogs.msdn.c
Hi,
I think is that correct. According to documentation, the compatibility mode
feature allows to write expressions that are not necessarily a number
value. In Mysql mode, you can write the expression LIMIT ,
and in the DB2 you can write OFFSET {ROW | ROWS}. However, the
limit expression is not
Hi Mike,
I would like to try to help you. So, can you provide schema/test data to
reproduce the cited situation?
Regards,
Fred
2015-03-09 10:24 GMT-03:00 Mike Goodwin :
> Hi,
>
> I have a query performance problem that has come, gone away and come back
> again.
>
> The issue seems to be when u
Hi,
You would need to use the server mode, or the automatic mixed mode.
Regards,
Thomas
On Saturday, January 31, 2015, Peter Cheung wrote:
> Hi
> I use java the start connect to h2 using jdbc. Because all threads in
> java are in one single process. So I can connect to the H2 concurrently.
Hi,
I agree this is incorrect. A patch is welcome.
However, could you explain what problem you want to solve?
Regards,
Thomas
On Mon, Mar 9, 2015 at 11:20 AM, Vojtěch Hordějčuk <
vojtech.hordej...@gmail.com
> wrote:
> update: the same behavior applies to OFFSET ?
>
> Dne pondělí 9. března 201
Hi,
Such questions are typically best asked at StackOverflow.com.
You would need a function that converts the timestamp to a value with a 5
minute delay.
Regards,
Thomas
On Wednesday, January 14, 2015, Hannes Hörting wrote:
> Hello!
>
> I write each minute in the databse:
> timestamp ; uid ;
Hi,
I don't know where it is converted to lowercase, so I'm afraid I can't help
here.
However, I would not use the column names "FROM" and "TO" (quoted or
unquoted).
Regards,
Thomas
On Sun, Mar 8, 2015 at 7:09 PM, wener > wrote:
> @Data
> @Accessors(chain = true, fluent = true)
> @Entity
> @T
Hi,
That sounds like a good idea! On the API side, I would probably make it a
key-value list, by extending the existing
table information_schema.settings. Otherwise the table will get a *lot* of
columns if we add other data. Of course, the problem of a key-value list is
that the datatype for both
Hi,
I'm not sure what features you want to add so I can't be certain, but I
think it is possible.
I suggest to read all the documentation about linked tables. Then, you may
want to read (part of) the implementation of that feature. It is probably
more complex for what you need.
Regards,
Thomas
Hi,
The "belongs_to_table" is not part of the official API. It is used
internally mark sequences as belonging to tables. The issue is, in your
case, that you forgot(?) to link the table to the sequence. It should be:
create table test(id identity);
script drop;
CREATE SEQUENCE PUBLIC.SYSTEM_SEQU
Hi,
Well the error message says timeout, are sure it is not a timeout?
Regards,
Thomas
On Tuesday, January 27, 2015, wrote:
> I am using the H2 1.4.185.
>
> The url is jdbc:h2:tcp://
> 127.0.0.1/activiti?TRACE_LEVEL_FILE=4;WRITE_DELAY=10;CACHE_SIZE=4096;MAX_COMPACT_TIME=5000
> ;
>
> ### Er
Hi,
I replied off list with the data. Thanks Fred.
I did a little bit of investigation myself. I could see in the optimizer
that for whatever reason it was not giving the plan which used the index a
better score. Actually the costs were all very similar (+/- 1%). I added in
an extra heuristic whi
Hi,
One way to debug is to look at the stack trace, and then set a breakpoint
there, and then check what the parameters / variables are.
Regards,
Thomas
On Thu, Mar 5, 2015 at 11:46 AM, Max Telfer wrote:
> Hi again
>
> I get the same result with any browser - Chrome, IE, Firefox. This used
Thanks very much Thomas!
2015-03-09 18:14 GMT+01:00 Thomas Mueller :
> Hi,
>
> The "belongs_to_table" is not part of the official API. It is used
> internally mark sequences as belonging to tables. The issue is, in your
> case, that you forgot(?) to link the table to the sequence. It should be:
>
Dear Thomas,
you can find attached a groovy file with the error reproduced.
Just launch from your command line: $> groovy example.groovy
My current environment is: Groovy Version: 2.1.6 JVM: 1.8.0_05
Best,
Alessio
Il giorno domenica 1 marzo 2015 01:58:35 UTC-8, Thomas Mueller ha scritto:
>
>
Hi,
The "FileLock" is the mechanism to ensure only one process accesses the
database file. If there is a "concurrent update" exception, the problem is
most likely a file system that is not supported by this mechanism.
I think operating system file level locking is required, and the FileLock
mecha
19 matches
Mail list logo