@nigel
i am used select * from .. Where,,,,, If the table have little field (3-6 field).
I am Confuse :
  1. start Running feb '2008 - November 2008 = Well Done, Still fast.
  2. Start Desember 2008 - Janary 2009       = Slow if not Remove data and backup & restore. after that,.. fast
  3. until now.... need 2 week to remove record transaction to table backup and must be run backup restore..
Why ???

--- On Mon, 4/27/09, Nigel Weeks <nweeks@...> wrote:

From: Nigel Weeks <nweeks@...>
Subject: RE: [firebird-support] Still About performance and running slow
To: firebird-support@yahoogroups.com
Date: Monday, April 27, 2009, 2:50 AM








Hmm...

You aren't doing any highly CPU intensive queries like:
'select count(*) from table' for paging through records?

i.e., Page 4 of 300 records with 50 records per page?

(Terribly sorry if this answer has already been provided...)

Nige.

> For All....
> I am sorry... i am wrong sent the database statistic. and this database
> statistic and Ibconsole after my application Slow down....
> I am Confuse because this application start running at feb 2008. and until
> Desember 2008 starting slow down .. and try to move record to database backup
> and run Backup and restore, well done. But start february 2009 every 1 month
> must be remove data transaction to backup database and must be backup restore
> for more faster on my application. but this month, 2 week running my application
> slow down again. Need correcting on my database server and parameter to make the
> best performance.
> Spec Server :
>  Window server 2000
>   interbase 6.0
>   Intel P IV xeon
>   2 gb RAM
>  HD scsi 100 gb
>  40 Client Computer to access database
>  
> IbConfig:
> #V4_LOCK_MEM_ SIZE  98304
> #V4_LOCK_SEM_ COUNT  32
> #V4_LOCK_SIGNAL   16
> #V4_EVENT_MEM_ SIZE  32768
> #DATABASE_CACHE_ PAGES  10000
> #SERVER_PRIORITY_ CLASS  1
> #SERVER_CLIENT_ MAPPING  4096
> #SERVER_WORKING_ SIZE_MIN 0
> #SERVER_WORKING_ SIZE_MAX 0
> #V4_LOCK_GRANT_ ORDER  1
> #ANY_LOCK_MEM_ SIZE  98304
> #ANY_LOCK_SEM_ COUNT  32
> #ANY_LOCK_SIGNAL   16
> #ANY_EVENT_MEM_ SIZE  32768
> #CPU_AFFINITY   1
>
> Database Statistic:
> Database "c:\data\myxpossdb2 50409.gdb"
>  
> Database header page information:
>  Flags   0
>  Checksum  12345
>  Generation  8840430
>  Page size  8192
>  ODS version  10.1
>  Oldest transaction 8840412
>  Oldest active  8840413
>  Oldest snapshot  8840413
>  Next transaction 8840419
>  Bumped transaction 1
>  Sequence number  0
>  Next attachment ID 0
>  Implementation ID 16
>  Shadow count  0
>  Page buffers  0
>  Next header page 0
>  Database dialect 3
>  Creation date  Apr 4, 2009 7:52:21
>  Attributes  force write
>  
>     Variable header data:
>  Sweep interval:  20000
>  *END*
>  
>  
> Database file sequence:
> File c:\data\myxpossdb25 0409.gdb is the only file
>  
> Database log page information:
>  Creation date
>  Log flags: 2
>   No write ahead log
>  
>  Next log page: 0
>  
>     Variable log data:
>  Control Point 1:
>   File name: 
>   Partition offset: 0  Seqno: 0  Offset: 0 
>  Control Point 2:
>   File name: 
>   Partition offset: 0  Seqno: 0  Offset: 0 
>  Current File:
>   File name:  
>   Partition offset: 0  Seqno: 0  Offset: 0 
>  *END*
>  
> Analyzing database pages ...
> MYSS (179)
>     Primary pointer page: 319, Index root page: 320
>     Data pages: 744, data page slots: 1599, average fill: 49%
>     Fill distribution:
>   0 - 19% = 254
>  20 - 39% = 39
>  40 - 59% = 27
>  60 - 79% = 263
>  80 - 99% = 161
>  
>     Index MYSS1 (1)
>  Depth: 2, leaf buckets: 11, nodes: 11558
>  Average data length: 0.00, total dup: 10313, max dup: 172
>  Fill distribution:
>       0 - 19% = 0
>      20 - 39% = 0
>      40 - 59% = 3
>      60 - 79% = 0
>      80 - 99% = 8
>  
>     Index MYSS2 (0)
>  Depth: 2, leaf buckets: 15, nodes: 11541
>  Average data length: 0.00, total dup: 11517, max dup: 1916
>  Fill distribution:
>       0 - 19% = 1
>      20 - 39% = 1
>      40 - 59% = 8
>      60 - 79% = 1
>      80 - 99% = 4
>  
>     Index MYSS3 (3)
>  Depth: 2, leaf buckets: 13, nodes: 11558
>  Average data length: 1.00, total dup: 9792, max dup: 172
>  Fill distribution:
>       0 - 19% = 1
>      20 - 39% = 1
>      40 - 59% = 2
>      60 - 79% = 1
>      80 - 99% = 8
>
> ASAP, Hope Solution/idea/ correcting on my database server.
> Thanks....
>
> --- On Fri, 4/24/09, Leyne, Sean <Sean@BroadViewSoftw are.com> wrote:
>
> From: Leyne, Sean <Sean@BroadViewSoftw are.com>
> Subject: RE: [firebird-support] Still About performance and running slow
> To: firebird-support@ yahoogroups. com
> Date: Friday, April 24, 2009, 5:53 PM
>
>
>
>
>
>
>
>
> Douglas,
>
>> On Fri, Apr 24, 2009 at 5:22 PM, Leyne, Sean
> <Sean@broadviewsoft w are.com>
>> wrote:
>> >> Oldest transaction 434
>> >> Oldest active 8366
>> >> Oldest snapshot 8366
>> >
>> > You have a database transaction that is not committing!
>>
>> The difference between Oldest Transaction and Oldest Active is the
>> number o transactions that need sweeping.
>> When transactions are left uncommited, it's the difference between
>> Oldest Active and Next that increases.
>
> You are correct, my bad!
>
> I was too quick in reviewing the details... ;-[
>
> Sean
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> [Non-text portions of this message have been removed]
>
>

--
Nigel Weeks

Prism9 Technology
e: nweeks@prism9. com.au
m: 0408 133 738
w:http://prism9. com

fb:http://facebook. com/profile. php?id=604592926



















[Non-text portions of this message have been removed]