Re: Config for OLTP system

On Wed, 10 Aug 2005 22:38:20 +0100, "Neil Truby" <neil.truby@ardenta.com> wrote: Opinions listed below . . . >IDS 10.0 v3 on Solaris 9. > >I'd be grateful for any comments on this OLTP system onconfig about to go >live on a SunFire v480 with 2 processors and 8GBytes of RAM. I have only 6 >extrenal disks fo the 60GByte database server. I've striped the database >across 3 of them with a 32k interlace then mirrored this stripe in a RAID >0+1 arrangement. > >Somehow it seems a little sluggish, though we've done no detailed >comparative testing .... > >thanks >Neil > >ROOTSIZE 512000 # Size of root dbspace (Kbytes)>PHYSDBS physdbs # Location (dbspace) of physical log >PHYSFILE 511894 # Physical log file size (Kbytes) Physical and logical logs in rootdbs??? ># Logical Log Configuration >LOGFILES 819 # Number of logical log files >LOGSIZE 1000 # Logical log size (Kbytes) Seems small . . . . .how many logs are backed up in an hour??? >LOG_BACKUP_MODE MANUAL # Logical log backup mode (MANUAL, CONT) ># Security ># DBCREATE_PERMISSION: ># By default any user can create a database. Uncomment DBCREATE_PERMISSON to ># limit database creation to a specific user. Add a new DBCREATE_PERMISSION ># line for each permitted user. >#DBCREATE_PERMISSION informix I like this one . . . .. 8-) >NETTYPE # Configure poll thread(s) for nettype Curious what the defaults are for this . . . . >DEADLOCK_TIMEOUT 60 # Max time to wait of lock in distributed env. >RESIDENT 1 # Forced residency flag (Yes = 1, No = 0) Is -1 a valid >MULTIPROCESSOR 1 # 0 for single-processor, 1 for>multi-processor >NUMCPUVPS 4 # Number of user (cpu) vps May be right for Sun . . . . >LOCKS 20000 # Maximum number of locks Very well-behaved applications? Seems low . . .. .what's ovlock show in onstat -p?? >NUMAIOVPS 3 # Number of IO vps Any KAIO possibilities for Sun?? >PHYSBUFF 10240 # Physical log buffer size (Kbytes) >LOGBUFF 1024 # Logical log buffer size (Kbytes) Flush a logical log buffer and then back it up to tape? I'd be curious of 'onstat -l' output and log percentages when flushed. >CLEANERS 33 # Number of buffer cleaner processes >SHMVIRTSIZE 327680 # initial virtual shared memory segment size >SHMADD 32768 # Size of new shared memory segments No additional segments?? That's not good in the HP world, may be different in the Sun world. >DYNAMIC_LOGS 2 >LTXHWM 70 >LTXEHWM 80 Personal preference here . . . . . >BAR_MAX_BACKUP 0 Doesn't this open onbar up wide open when it comes backup time? Number of dbspaces by default? >RA_PAGES # Number of pages to attempt to read ahead >RA_THRESHOLD # Number of pages left before next group What are the defaults for 10? >OPTCOMPIND 2 # To hint the optimizerWhat about 0 for OLTP? ># The following parameters are related to the buffer pool >BUFFERPOOL >size=2K,buffers=1024000,lrus=33,lru_min_dirty=2.000000,lru_max_dirty=10.000000> Max of 6M per LRU to flush? What are the checkpoint durations? Just a few ideas . . . . . FWIW JWC