Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Mailing Lists -> Oracle-L -> Re: Random long times to connect using BEQ adapter
Rjamya,
is it RAC by any chance? i have had weird problems on AIX 5.3/RAC when one of the nodes has a communication problem. ensure that there are no comms issues at all.
Some in the dark abyss of my memory, i recollect that Oracle had a way of assigning file numbers to the shared libraries it opens when you connect, just can remember.
may be an SR to understand the file number part??? it may also be that the file permission is not as per the requirement. truss would report an error if the file is not found. is this some shared OH config? NFS?
not to discourage you, but i had the same nasty problem on a Tru64 GS1200 machine a couple of years ago. the application was crappy and connecting/disconnecting a million times a second. we used tonnes of listeners, BEQ, nothing helped.
Oracle just could not do anything even with a P1 TAR.
hope this helps,
anand
On 20/11/06, Stefan Knecht <knecht.stefan_at_gmail.com> wrote:
>
> To figure out what's behind fd10, you could try to look at lsof (might
> have to get your SE to install it first, on aix it's not there per default
> afaik).
>
> Then grep for your sqlplus' PID - IF it can write to it, it has to have it
> open.
>
> Stefan
>
> On 11/20/06, rjamya <rjamya_at_gmail.com> wrote:
> >
> > After a cpu upgrade, local connections are taking exceeding long times
> > (between 7 to as much as 40 seconds) using sqlplus. Once connected, it
> > is blazingly fast to run the script.
> >
> > I did a "truss -aefdo" on the sqlplus session and following is a
> > snipped that shows delay. For the life of me I can't find out what the
> > heck fd 10 is.
> >
> > 1597872: 3314117: 0.2855: kfcntl(4, F_SETFD, 0x0000000000000001)
> > = 0
> > 1597872: 3314117: 0.2857: lseek(4, 0, 0) = 0
> > 1597872: 3314117: kread(4, "1513 "011303\t\t\0\0\0\0".., 256) = 256
> > 1597872: 3314117: 0.2861: close(3) = 0
> > 1597872: 3314117: 0.2862: close(4) = 0
> > 1597872: 3314117: 0.2864:
> > open("/usr/opt/oracle/9204/sqlplus/mesg/sp1us.msb", O_RDONLY) = 3
> > 1597872: 3314117: 0.2865: kfcntl(3, F_SETFD, 0x0000000000000001)
> > = 0
> > 1597872: 3314117: 0.2867: lseek(3, 0, 0) = 0
> > 1597872: 3314117: kread(3, "1513 "011303\t\t\0\0\0\0".., 256) = 256
> > 1597872: 3314117: 0.2870:
> > open("/usr/opt/oracle/9204/sqlplus/mesg/sp2us.msb", O_RDONLY) = 4
> > 1597872: 3314117: 0.2872 : kfcntl(4, F_SETFD, 0x0000000000000001)
> > = 0
> > 1597872: 3314117: 0.2874: lseek(4, 0, 0) = 0
> > 1597872: 3314117: kread(4, "1513 "011303\t\t\0\0\0\0".., 256) = 256
> > 1597872: 3314117: kwrite(9, "0186\0\006\0\0\0\0\011 i".., 390) = 390
> > 1597872: 3314117: 2.2880: kread(10, "\0
> > Ì\0\006\0\0\0\0\004\0".., 2064) (sleeping...)
> > 1597872: 3314117: kread(10, "02 %\0\006\0\0\0\0\0101D".., 2064) = 549
> > 1597872: 3314117: 12.1737 : lseek(8, 1024, 0) = 1024
> > 1597872: 3314117: kread(8, "\018\0 $\0 7\0 @\0 J\0 V".., 512) = 512
> > 1597872: 3314117: 12.1741: lseek(8, 47104, 0) = 47104
> > 1597872: 3314117: kread(8, "\0\f05 ^\0\0\0 P05 x\0\0".., 512) = 512
> > 1597872: 3314117: kwrite(9, "\0 4\0\006\0\0\0\0\011 i".., 52) = 52
> > 1597872: 3314117: kread(10, "\014\0\006\0\0\0\0\0\b\0".., 2064) = 20
> > 1597872: 3314117: kwrite(9, "\0 Ý\0\006\0\0\0\0\00303".., 221) = 221
> > 1597872: 3314117: kread(10, "\090\0\006\0\0\0\0\004\0".., 2064) = 144
> > 1597872: 3314117: kwrite(9, "\0 Â\0\006\0\0\0\0\003 G".., 194) = 194
> > 1597872: 3314117: kread(10, "\09B\0\006\0\0\0\0\0\b\0".., 2064) = 155
> > 1597872: 3314117: kwrite(9, "\015\0\006\0\0\0\0\00305".., 21) = 21
> > 1597872: 3314117: kread(10, "\0 ª\0\006\0\0\0\0\004\0".., 2064) = 170
> > 1597872: 3314117: kwrite(9, "\011\0\006\0\0\0\0\003\b".., 17) = 17
> > 1597872: 3314117: kread(10, "\00F\0\006\0\0\0\0\0\t\0".., 2064) = 15
> > 1597872: 3314117: kwrite(9, "01 7\0\006\0\0\0\0\003 ^".., 311) = 311
> > 1597872: 3314117: kread(10, "\0 Þ\0\006\0\0\0\0\0\v01".., 2064) = 222
> > 1597872: 3314117: kwrite(9, "01 6\0\006\0\0\0\0\011 i".., 310) = 310
> >
> > see, in the third column the time reported (delta due to -d switch)
> > jumps from 2.288 to 12.1737? In the whole file, there is no reference
> > to a open or kopen for fd 10.
> >
> > Any other ideas to identify why this slowness? Any other way to trace?
> > Running oracle code is fine, it is just the slowness when I issue a
> > "conn user/pass" in sqlplus. This is aix 5.3, oracle 9204.
> >
> > TIA
> > Rjamya
> > ----------------------------------------------
> > Got RAC?
> > --
> > http://www.freelists.org/webpage/oracle-l
> >
> >
> >
>
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Nov 20 2006 - 22:58:58 CST