|
testcase
normal
connection
pooled
connection
%improvement
connect
and disconnect
1000 times
226 sec
154 sec
31%
connect,
insert one row,
disconnect
1000 times
309 sec
178 sec
42%
this results are very promising. first we see 30% improvement of time to connect
to db. what is more important we can see 40% less time on repeating the same
simple statement after establishing each connection. the only explanation is
that pooled connection can reuse already parsed statement. that is good news if
we think about scalability.
you may be also interested in
oracle
official white-paper on drcp.
at the end i would like to thank
christopher
jones as he was involved in drcp development. i've mailed with him few times
last year about problems with php-oracle connectivity and lack of connection
pooling.
cheers, pawe?
ibm’s debut of its homegrown open source version of openoffice without e-mail or collaboration features is not surprising but nevertheless disappointing.
on the one hand, big blue’s recent endorsement and support and services plan for openoffice — an acknowledgement of the obsolescence of smartsuite and workplace — offers new hope for the struggling open [...]
the ibm symphony tools the company is now giving users free is aimed at nothing less than jump-starting a social revolution in the workplace.
9.2.0.8 patching on vmswell, at least one installation went well and i suppose if i had to pick one, it would have been the big, bad alpha/vms production server. the patching was a little interesting. i felt like i'd been transported back 6 or 7 years with the plain text warnings, ...
for internal reasons, this patchset will not be installed by the oracle
universal installer.
instead, the patchset will be installed in a manner
similar to regular patches on oracle openvms.
reviewing output files and relinking the binaries, but i've obviously found my correct place in the industry, because it all felt quite comforting seriously, it shows i can do the business when working with decent kit, proper planning and business pressures!
back to the 11g, then. i thought it had gone well but when i checked a little more carefully, the instance wasn't up. by now i realised it must be me, right? how about re-starting it and seeing what the problem is ...
sql> startuporacle instance started.
total system global area 1540046848 bytesfixed size
1300100 bytesvariable size
922749308 bytesdatabase buffers
603979776 bytesredo buffers
12017664 bytesdatabase mounted.ora-00333: redo log read error block 41823 count 8192eh? oh, dear, i vaguely remember having some disk problems months ago that mysteriously disappeared. mmm, better check that in /var/log/messages.sep 18 18:32:47 isp4400 kernel: info fld=0x3c7fc18, current sdc: sense key medium errorsep 18 18:32:47 isp4400 kernel: additional sense: unrecovered read errorsep 18 18:32:47 isp4400 kernel: end_request: i/o error, dev sdc, sector 63437824sep 18 18:32:58 isp4400 kernel: scsi0: error on channel 0, id 2, lun 0, cdb: read (10) 00 03
fantastic! sigh .... back to the drawing board
p.s. and this isn't even the 'one to go'. that's the oem grid control agent on 10g/vista. i think i'll concentrate on the 11g server for now.
(责任编辑：admin) |
