logs archiveIRC Archive / Freenode / #oracle / 2010 / June / 2 / 1
zasz
RJaway:
RJarett:
RJarett
howdy
zasz
it was the maxbytes on the file :(
i didn't realize you had to manually adjust that
RJarett
figured something like that
zasz
i sent out an email and my co worker looked at me like i was a retard
the other one who is older and more experience was like yea you'll run into silly stuff like that when you're learning oracle, don't worry about it
felt like a fu*king retard though
heh
after all that sh*t i did
stede
is it possible to create a job that executes an stored procedure every N minutes, but only if the previous call already succeses ?
usn_work
good afternoon
is there a way to have only pre-compiled PL/SQL on a server, in order to hide the source code? (I already know wrapping, but that's not hte same.
)
I know oracle to import their internal packages in binary, but I don't know how they do it
ech0s7
hi
i have created a directory with this command: create directory ReportDump as '/home/oracle/ReportDump';
sqlplus says OK. but in /home/oracle/ i haven't ReportDump
usn_work
create it
the directory object is just a casing
         

ech0s7
usn_work: ok, and after if i use expdp with this directory, where can i find export dump file ?
usn_work
in /home/oracle/ReportDump
if you created it
you need the database's directory object to be able to reference a directory on OS level
but if it does not exist on OS level, the dir object is just a null pointer if you want
ech0s7
ok thanks usn_work
usn_work
np
kaylee_work
morning
usn_work
evening ;)
kaylee_work
I've got a table with a number of events in it each with a start and end date... what would be the best way to get a row for every day in a specific date range, and the count for all events that start on that day, and then another count for all events that started that day that also ended that day?
I've been able to get a row for each day that's got events starting... but don't get a row for any days that have no events
http://pastebin.com/iAHnYCz4
usn_work
kaylee_work, won't develop youir SQL, but maybe something called "inline select": SELECT field1, field2, (select ....), field3 FROM table WHERE <condition> GROUP BY ... ORDER BY ...
the inliner is the one in brackets
but this is really sh*tty ;)
kaylee_work
interesting
(Action) reads
generally speaking, would you recommend going down that route? or is it better just to select all the data in that date range and have my app massage the data, instead of trying to get SQL to do it?
usn_work
this decision is something you have to take, based on benchmarking it, (or calculating the costs)
may be good this way o another
kaylee_work
hum... I'm kinda leaning toward the simple "I can write it faster not in sql" route :D
do appreciate the info though
usn_work
writing it quick does not necessarily mean "running quick"
:)
kaylee_work
I'd guess that one simple select would process faster than a complicated one with a bunch of sub-selects
usn_work
this may be true, but there can be a downside as well
kaylee_work
of course there's also the maintainability issue
and the ever popular KISS
usn_work
this is third line IMHO
just test the runtime
should be priority, and BAAG
Battle against any guess
kaylee_work
lol
usn_work
I'm doing performance stuff all day long, can't afford believing in nice-sounding theories. I've got 200 developers here, each with a nice new one
:)
kaylee_work
lol
mgolisch
what was it again one has to do to have sheduler jobs executed as another user than nobody?
i think i had to setuid some file but i forgot which one
         

b_52light
hi
plz ppl is there a similar element to xmlroot but without doctype
krb_hnd
Hi all
how to use impdb in parallel mode ?
could anyone here share some knowledge?
hali
krb_hnd: parallel=N
but requires enterprise edition
(where n is the parallelism obviously, not an 'N')
krb_hnd
can I use impdb
?
hali
yes
krb_hnd
is this command correct "impdb username/pass parallel=4" then where to put the file name?
now I have 4 files parallel_01 ...04.dmp
I have exported all the data to oracle directory
but I got problem when tried to tranfer file to other machine
it takes more then 4 hours
to transfer files
hali
copying files? not really an oracle problem
you can specify a list of files
in dumpfile=
dumpfile=file1.dmp,file2.dmp etc
krb_hnd
ok
Ramen
how are you transfering?
krb_hnd
hali: currently I'm trying to sync between to db in different machine
hali : can you suggest me the best tools to do that
hali
scp?
krb_hnd
Ramen : I'm planing to used nfs
hali
make sure you have the correct nfs mount options
krb_hnd
data is +- 10g
hali
oracle is dog slow with the incorrect options, if even works
10gb should be copied in no time
Ramen
maybe you can upgrade the network O.o
krb_hnd
it takes more than 4 hours to complete
hali
is use rw,bg,hard,nointr,tcp,vers=3,timeo=300,rsize=32768,wsize=32768,actimeo=0
i've seen rman backups to incorrectly configured NFS take ages and ages
krb_hnd
?
Ramen
you guys at gigbit yet?
hali
krb_hnd: in fstab
i use netapp snapmirror/flexclones to keep our test/dev/qa db's up in sync
krb_hnd
is it free
hali
no, it's a SAN feature that starts at about 100k USD
krb_hnd
oo ok
then I have to proceed the normal method
expdb and impdb
hali
check nfs mount options first
krb_hnd
in fstab server
hali
on the client (oracle master server?)
to see if you can speed up the NFS access, wire speed should be what you can expect from decent hardware on gigabit these days
krb_hnd
I insert : servername:/path rw=client_ip:client_ip
is this correct
hali
that looks like an NFS server setting, not the client mount settings
i presume you have the filesystem in /etc/fstab on the NFS client
krb_hnd
yes
hali
stick those options above in instead of "default"
krb_hnd
that one is for server
hali
and remount
krb_hnd
how about client hali
first I start the nfs.client
if this method failed
hali
?
umount and mount and you are done
krb_hnd
then I'll proceed my second back up which is impdb via network
hali
are you using vms or something? :)
yes
krb_hnd
nope
I think I need to add something in client right?
b_52light
plz ppl is there a similar element to xmlroot but without doctype
or if ii have an xml element like <?xml version="1.0"?> <ROW> <ENAME>John</ENAME></ROW> , how could i remove the doctye
krb_hnd
ok
got to go now
anyway
thanks for the help
really appreciate it
esp hali
bye
rato
is there any book about "UCM" ?
what's the relation between UCM and Stellent Content Server ?
Ramen
Number(3,1) means a number 3 long with 1 decimal place right? like 12.3 and 123.4 is not valid?
sorry i'm so noob at this :(
rato
Ramen it is. Number(3,1) means a number like this 12.3
\part
dddh
_jst: hm
SeveredCross
Hi guys, I seem to be having some trouble connecting to an Oracle server through a VPN. I've got SQLplus set up on my local box, and the VPN is connected, but when I try to do sqlplus user/pass@//server-ip/service-name, I get ORA-12514: TNS:listener does not currently know of service requested in connect descriptor.
DoubleString
„yo
SeveredCross
However, the server is live, because SSHing in to the machine and running sqlplus user/pass@SID works just fine.
Er, sorry, with the SID exported in ORACLE_SID.
Never mind, I was specifying the wrong service name. on my local box.
« prev next »