logs archiveIRC Archive / Freenode / #oracle / 2009 / December / 2 / 1
tunke
hi all
Oracle use KB =1000 or KB = 1024?
razym
1024 if i'm not mistaken
-_-
bicchi
how do i convert a linux/unix time stamp format to a full date and time format? let me explain.
in linux if i type: $ date +"%s
i get 1259723390
i wish in SQL to translate the 1259723390 to 12/1/2009 22:10:37
i have a table that the date and time is formated as time stamp and i wish to display it the as MM/DD/YY HH:MM:SS
Dormir
use a format like this: date +"%m/%d/%Y %H:%M:%S"
bicchi
no i mean in SQL. this is an SQL question not a linux problem. that was just an example.
Dormir
yes, understood
if you can't change the format from *nix, you need to figure out what %s gives you
I think it's seconds since a certain date
then it's just a case of date math
bicchi
i was just thinking that there might be a function built in or something.
Dormir
not that I remember, but I haven't looked in awhile
Kaell
i'd like to select "foo"||"bar" from sometable; is this possible?
ideally, just using the raw column name of foobar, but I can alias it to "foobar" if necessary
         

Dormir
look up dynamic sql
in the docs
Kaell
hmm, thanks... looks like that makes the stored proc not cacheable though... so i'll just use lots of IF/THEN blocks instead... cheers
HSXN
can anyone tell me how to configure the Microsoft LoopBack Adapter to be the primary network adapter on the system.
its telling me that i have a warning on network configuration because im dhcp assigned. (during installation, checking prerequisites)
HSXN__
that was weird, my connection to irc just failed. anyways, when connecting to the enterprise manager database control, at the local url, what would the default username be?
cmug
how do I check if I have _b_tree_bitmap_plans set to true or false?
TheBonsai
not sure, maybe v$parameter lists hidded parameters, or try ''show parameter _b;''
cmug
I'll take a look, thanks
the show parameter doesn't work
I can see it in v$parameter, but I do not know is it there by default (I have set already). I'll look at a virgin db
yeah it is not there if it has not been set
I wonder what is the default value, True probably
TheBonsai
show parameter can't list it?
(i have no db here to check, sorry)
cmug
after it has been set it does
but before no
(which makes sense, since it is a hidden parameter)
I guess show parameter queries v$parameter and shows what is there
TheBonsai
hey....
mysql admins:
<kbp> I need help!!! I accidentally drop database in phpmyadmin instead of drop table... is there anyway to recover it in ubuntu?
(##linux)
if these are the alternatives, my job is safe.
jmnemonic
hi, i'm trying to imitate the LIMIT sentence from MySQL in Oracle and I can't
i mean
TheBonsai
where rownum < ...
jmnemonic
yeah, that works
I've tried
but
for example
if I wanna get the first ten rows
where rownum < 10
but if I wanna get the rows between 10 and 20 ?
TheBonsai
where rownum between 10 and 20 ...?
jmnemonic
it doesn't work
TheBonsai
i have no db here to test that, still at home
jmnemonic
alrigth
TheBonsai
and what if you AND 2 expressions?
did you test that?
jmnemonic
where rownum > 10 and rownum <= 20
TheBonsai
yea
         

jmnemonic
it didn't work
TheBonsai
then check the asktom article about it, i just have no link :)
jmnemonic
ok
i'll google it
thanks
TheBonsai
well, maybe a subexpression
select * from (select ..... rownum as foo) where foo between.....;
such constructs
jmnemonic
i'm trying to do something like this but it doesn't works
select * from (select * from tablename) where rownums <= 10
rownum*
corpsicle
hah, trying to install 11g on an old linux virtual machine, all the prerequisite checks failed >_<
_all_
DiscordianUK
That's why there's a --ignoreSysPrereqs flag
corpsicle
well one of them is that /tmp is too small =/
that has to be fixed afaik
DiscordianUK
Yes that would need fixing
corpsicle
swapspace is 1.94 and not 2.01 >_<
oh noes
DiscordianUK
Sounds familiar, just add a small swap file and lose it after
corpsicle
i could probably ignore that one though
i doubt its going to crash because its missing 0.06 gb swap
however the tmp thing is annoying
DiscordianUK
It won't crash , if it's using 2Gb swap you've got other worries
corpsicle
heh
DiscordianUK
Like it b e i n g v e r y s l o w
jmnemonic
About numrow and trying to imitate the limit function
I found a solution using adodb
and it worked
:)
the connection object has a method called Selectlimit(sql, limit, offset)
and that was all
well, i forgot to mention i was working with oracle and php scripts
corpsicle
uh wtf, i just pressed the "check again" button and now every check has "n/a" as expected and "n/a" as actual
lol
i need to start this installer over
what an annoying bug
DiscordianUK
The Oracle Installer isn't fun
corpsicle
uh, fixed the permissions on /tmp, restarted the installer, and its still giving me all n/a values
"ignore all" >_<
DiscordianUK
The old text mode installer on Oracle 7 took about 2 days to do a full install from 6 QIC tapes
corpsicle
lol
DiscordianUK
Mind you that was 10 years ago on a machine that cos £1 milliion
corpsicle
>_<
DiscordianUK
s/ii/i/
corpsicle
gah
its stuck at "execute root scripts"
i didnt get a popup
but i ran the scripts
and now i cant continue >_<
hello again
now i have 11g installed
but i also have 10g installed on the same machine
DiscordianUK
uhuh
corpsicle
when i try to start the listener i set up for 11g lsnrctl only shows the 10g listeners
and i know im using the correct binary
and the ORACLE* vars are all set to point to the 11g installation
DiscordianUK
Uhuh
corpsicle
ideas ? =P
DiscordianUK
You'll need to update your listeners.ora to include both 10g and 11g instances
corpsicle
why both ?
when i do lsnrctl status theres this also : Listener Parameter File /opt/oracle/product/10.2.0/network/admin/listener.ora
Listener Log File /opt/oracle/product/10.2.0/network/log/listener.log
why is it using the 10.2.0 path to check these files ?
DiscordianUK
you can only have one listener daemon on the standard tcp port
corpsicle
when im running the binary from the 11.2.0 path
uhu
DiscordianUK
It does what you tell it to
corpsicle
so it connects to the PORT ?
it doesnt care what path or variables
interesting
but i could run 10g and 11g listeners on the same port ?
DiscordianUK
port 1521
use the 11G tns listener
corpsicle
LSNRCTL for Linux: Version 11.2.0.1.0 - Production on 02-DEC-2009 12:48:15
Version TNSLSNR for Linux: Version 10.2.0.1.0 - Production
o_O
does that say the binary im running is 11g but the daemon its connecting to is 10g
DiscordianUK
and modify the listener.ora to include both 10g and 11g instances
corpsicle
wow, now i get it
just stopping and starting got me onto the 11g listener.ora config
right, ill put all of them in the 11g config
hali
corpsicle: you could run the 11g listener serving both instances, not running two listener process on the same port
TheBonsai
you can't have two processes listening on the very same socket
corpsicle
yes
TheBonsai
this "magic" is not really oracle specific
corpsicle
no
Franek_Kimono
how are you gentleman!
TheBonsai
m00
(Action) begins to like 11.2 grid infra
phimic
hey all
still got an problem with my oracle-rac on vmware esx using shared vmdk as raw device, the cluster runs ok for few minutes then i get the following error http://pastebin.com/d183199c2
mator
phimic, isn't it the first message you posted a week ago describing the same problem ?
phimic
mator: yes, i have opened a service request on novell and vmware but the dont know the reason
mator
phimic, did you tried "disk.locking = true" in config file ?
phimic
mator: i tried disk.locking="false"
mator
k
phimic
accoring to all blogs i read
sorry according
mator
phimic, its actually a scsi reservation thing (scsi-3), one node should put a lock on a shared disk, all other nodes should see that lock is set
when the lock can't be set or read, cluster member will reboot (to preserve data integrity)
phimic
i will try to add more voting disks
maybe it will be solved
mator
vcs (veritas cluster server) requires at least 3 reservation disks
I don't know how much does oracle require
also this disks should support scsi-3 reservation feature, so you should double check that it is work for your controller and disks
phimic
mator: do you know how i can check this in linux?
mator
phimic, better check with your vmware, should be in docs...
too bad i'm missed all the fun with vmware esx
got an email yesterday from local EMC branch HR girl
we're discussing me as a possible EMC engineer....
i told them that i don't have EMC products expirience, only HDS
phimic
mator: i found that "virtual machines support only SCSI-2 reservation protocol and not SCSI-3"
on this blog http://dsumsky.blogspot.com/2007/11/i-need-platform-for-testing-of-clusters.html
mator
phimic, does your vmware defined shared disks have its own controller (i.e. scsi 1:1) ?
phimic
mator: yes
i created a new controller scsi1:0 and all shared disk are on 1:1, 1:2, ... adapter typ LSI and shared is virtual
mator
http://www.oracle-base.com/articles/10g/images/10gRACLinux/32-VirtualMachineSettingsDiskAdvanced.jpg
the picture is example for shared disk in vmware
you could probably compare your configuration with this one http://www.oracle-base.com/articles/10g/OracleDB10gR2RACInstallationOnCentos4UsingVMware.php
phimic
mator: that is exactly what i did
mator
phimic, can you please port somewhere "esxcfg-info | grep -i scsi" ?
phimic
mator: i pasted it here http://pastebin.com/d33aee001
mator
phimic, and what VMkernel logs contain for that rebooted machine?
phimic
mator: i did not find anything interesting http://pastebin.com/d68ea7599
mator
phimic, seems i can't help... though I have 3 links for you
http://tinyurl.com/ykj62dv
http://tinyurl.com/ylkzcc8
http://tinyurl.com/ykycsff
phimic
mator: ok thank you will look at those links
TomJ
I've got an ASM instance that hung on startup. it paused on the line: "ASM Cache 25165824 bytes" for 10 minutes. then we control-C, and tried to shutdown, but it says "ORA-01089: immediate shutdown in progress - no operations are permitted"
there are two ASM processes running:
oracle 1104 1 0 19:24:51 ? 0:00 asm_rbal_+ASM
oracle 2034 2011 0 19:35:01 ? 0:00 oracle+ASM (DESCRIPTION=(LOCAL=YES)(ADDRESS=(PROTOCOL=beq)))
shall I kill them manually? any particular order?
ah, that LOCAL=YEs was sqlplus
so just asm_rbal_+ASM is running
i guess i have no choice but to kill it?
btw all this happened after the box rebooted itself with the database mounted
« prev 1 2 next »