Re: wrapping packages
Tanel Poder wrote: I checked, the wrap executable in 8.0.6 dist for solaris is about 3MB, but for 9.2 in Windows it's only about 40k. Perhaps you're not aware of the way executables compiled on your Solaris and Windows platforms. In detail, not. In general, yes. Ok, I checked, you're correct, wrap isn't only this 40kB executable, uses orancrypt9.dll (100kB) in Windows, this might be the one where encryption is done... It shouldn't be that hard to reverse engineer it. It's an extremely commendable plan... (a touch of irony here) :) I've dealt with disassembling before, back in old dos times (disassembling 4kB graphical intros and few viruses :). I don't think this is a hard job to do, it's just time consuming - it gets hard when the authors have planted debugger traps and various other tricks into the code that make the crackers life hard (or should I say interesting :) Tanel. Probably the ones who already have cracked the algorithm aren't spreading the knowledge - why should they anyway?! I suggest you to call Oracle legal and discuss this issue and your original plan of fixing it. :) -- Vladimir Begun The statements and opinions expressed here are my own and do not necessarily represent those of Oracle Corporation. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Vladimir Begun INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
Coming to think of it, is the 3Mb size is for stripped or unstripped executable?. I know 3Mb if stripped is not likely to come to 40K. Nevertheless the memory addressing for UNIX is diff. from Windoz. May be someone else in the list has a better knowledge of it. GovindanK -Original Message- From: Tanel PoderSent: 9/20/2003 2:44:48 PMTo: [EMAIL PROTECTED]Subject: Re: wrapping packagesI checked, the wrap executable in 8.0.6 dist for solaris is about 3MB, but for 9.2 in Windows it's only about 40k. It shouldn't be that hard to reverse engineer it. Probably the ones who already have cracked the algorithm aren't spreading the knowledge - why should they anyway?! Tanel. - Original Message - To: "Multiple recipients of list ORACLE-L" <[EMAIL PROTECTED]>Sent: Sunday, September 21, 2003 12:29 AM Hi Pete I must point out that there must be a unwrap, since the Oracle database can run the wrapped pl/sql code :-) It is based on trust in Oracle coopera! tion / development. Some times it would make since to write the code in c/c++ since it harder to revers. Pete Finnigan wrote: HiVery true, but if there was the wrap process wouldn't be much use as anyone could un wrap your code. But you are right the main reason to be cautious is to not delete your source code locally.kind regardsPeteIn article <[EMAIL PROTECTED]>, bhabani s pradhan <[EMAIL PROTECTED]> writes one caution:there is no unwrap cmd/exeRegards ___Get Your 10MB account for FREE at http://mail.arabia.com !Access MILLIONS of JOBS NOW!
Re: wrapping packages
Hi Did a nm -D wrap this gave heaps of symbols, so wrap is loading shared libs. This means revers engineering a lot of Oracle code :-( Govindan K wrote: Coming to think of it, is the 3Mb size is for stripped or unstripped executable?. I know 3Mb if stripped is not likely to come to 40K. Nevertheless the memory addressing for UNIX is diff. from Windoz. May be someone else in the list has a better knowledge of it. GovindanK -Original Message- *From: Tanel Poder* Sent: 9/20/2003 2:44:48 PM To: [EMAIL PROTECTED] Subject: Re: wrapping packages I checked, the wrap executable in 8.0.6 dist for solaris is about 3MB, but for 9.2 in Windows it's only about 40k. It shouldn't be that hard to reverse engineer it. Probably the ones who already have cracked the algorithm aren't spreading the knowledge - why should they anyway?! Tanel. - Original Message - To: Multiple recipients of list ORACLE-L Sent: Sunday, September 21, 2003 12:29 AM Hi Pete I must point out that there must be a unwrap, since the Oracle database can run the wrapped pl/sql code :-) It is based on trust in Oracle coopera! tion / development. Some times it would make since to write the code in c/c++ since it harder to revers. Pete Finnigan wrote: Hi Very true, but if there was the wrap process wouldn't be much use as anyone could un wrap your code. But you are right the main reason to be cautious is to not delete your source code locally. kind regards Pete In article , bhabani s pradhan writes one caution: there is no unwrap cmd/exe Regards ___ Get Your 10MB account for FREE at http://mail.arabia.com ! Access MILLIONS of JOBS *NOW*! http://ads.arabia.com/?SHT=text_email_english -- Peter Gram, Miracle A/S Phone : +45 2527 7107, Fax : +45 4466 8856, Home +45 3874 5696 mail : [EMAIL PROTECTED] - http://MiracleAS.dk Upcoming events: DatabaseForum 2003, Lalandia 2-4 October Visit http://miracleas.dk/events/DBF2003/invitation.html Miracle Master Class with Tom Kyte, 12-14 January 2004 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Peter Gram INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: weird
8.1.7.4 No AUM. Either it was something goofy in the database, or something clobbered some bytes as they were going from the host machine to my telnet session ... which afaik would also be wierd since TCP/IP is supposed to guarantee delivery. The terminal session is on a Windows box ... maybe that's it! -Original Message- Which version are you on? Just wondering if it might have something to do with some bug in automatic undo management? Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, September 19, 2003 6:09 PM No question here. Just something weird. This is a long-running insert with NO NONE ZERO ZIP ZILCH NADA commit. It makes me wonder if something weird is going on, or if I am overlooking something in the query. SQL select a.username,sum(b.used_ublk) x from v$session a, v$transaction b where a.taddr=b.addr group by a.username; USERNAMEX -- -- SYSTEM418 1 row selected. SQL / USERNAMEX -- -- SYSTEM893 1 row selected. SQL / USERNAMEX -- -- SYSTEM 2 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3181 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3204 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: weird
Just one transaction. Nothing else going on. -Original Message- Stephen, Was this the only running transaction? No background stuff going on? Jared -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: weird
Hi! After spending half of Saturday and Sunday digging around x$ktcxb and other interesting views I actually realized that the answer is way simpler :) TADDR column in v$session points to *current* transaction state object, that means if recursive transaction is needed for wrapping or extending rollback segment, TADDR points to this recursive transaction. When your query happens to select at the same time, it sees statistics for the small recursive transaction, not your big one. You should have been joining v$transaction.ses_addr with v$session.saddr instead of v$session.taddr with v$transaction.addr. (note there is a column recursive in v$transaction which allows to filter recursive transactions out) I tested it on 9.2.0.3 with AUM, but it should be the same in earlier versions. Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 4:04 PM 8.1.7.4 No AUM. Either it was something goofy in the database, or something clobbered some bytes as they were going from the host machine to my telnet session ... which afaik would also be wierd since TCP/IP is supposed to guarantee delivery. The terminal session is on a Windows box ... maybe that's it! -Original Message- Which version are you on? Just wondering if it might have something to do with some bug in automatic undo management? Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, September 19, 2003 6:09 PM No question here. Just something weird. This is a long-running insert with NO NONE ZERO ZIP ZILCH NADA commit. It makes me wonder if something weird is going on, or if I am overlooking something in the query. SQL select a.username,sum(b.used_ublk) x from v$session a, v$transaction b where a.taddr=b.addr group by a.username; USERNAMEX -- -- SYSTEM418 1 row selected. SQL / USERNAMEX -- -- SYSTEM893 1 row selected. SQL / USERNAMEX -- -- SYSTEM 2 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3181 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3204 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego,
RE: SGA Max size
In 9202, DB_BUBBER_CACHE, SHARED_POOL, LARGE_POOL and JAVA_POOL can be dynamically altered. But in 901, LARGE_POOL and JAVA POOL are static. If MAX SGA is less than 128MB then Oracle will use 4MB granule size to allocate/deallocate memory. For SGA greater than 128M, Oracle granule size is 16MB. -Original Message- To: Multiple recipients of list ORACLE-L Sent: 9/20/03 11:19 AM My understanding of SGA is SGA = x + y + z where x = (dbblksize*db_blk_buf OR db_cache_Size if 9i) y=shared_pool z=java pool, log_buffer If 9i oracle introduced SGA_MAX_SIZE; the sum of x+y+z can be SGA_MAX_SIZE; if so, which part of x/y/z expands when need arises. Thanks Quriyat ___ No banners. No pop-ups. No kidding. Introducing My Way - http://www.myway.com -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: quriyat INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). DISCLAIMER: This message is intended for the sole use of the individual to whom it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not the addressee you are hereby notified that you may not use, copy, disclose, or distribute to anyone the message or any information contained in the message. If you have received this message in error, please immediately advise the sender by reply email and delete this message. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: [EMAIL PROTECTED] INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
IIRC, 'wrap' does not actually encrypt the code. Rather, it simply does a precompile on it and then stores the pcode in the database. Jared On Sat, 2003-09-20 at 14:29, Peter Gram wrote: Hi Pete I must point out that there must be a unwrap, since the Oracle database can run the wrapped pl/sql code :-) It is based on trust in Oracle cooperation / development. Some times it would make since to write the code in c/c++ since it harder to revers. Pete Finnigan wrote: Hi Very true, but if there was the wrap process wouldn't be much use as anyone could un wrap your code. But you are right the main reason to be cautious is to not delete your source code locally. kind regards Pete In article [EMAIL PROTECTED], bhabani s pradhan [EMAIL PROTECTED] writes one caution: there is no unwrap cmd/exe Regards -- Peter Gram, Miracle A/S Phone : +45 2527 7107, Fax : +45 4466 8856, Home +45 3874 5696 mail : [EMAIL PROTECTED] - http://MiracleAS.dk Upcoming events: DatabaseForum 2003, Lalandia 2-4 October Visit http://miracleas.dk/events/DBF2003/invitation.html Miracle Master Class with Tom Kyte, 12-14 January 2004 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Peter Gram INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Jared Still INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: weird
Tanel, This is above and beyond the call of duty. Now you need to step away from the computer... Jared On Sun, 2003-09-21 at 08:29, Tanel Poder wrote: Hi! After spending half of Saturday and Sunday digging around x$ktcxb and other interesting views I actually realized that the answer is way simpler :) TADDR column in v$session points to *current* transaction state object, that means if recursive transaction is needed for wrapping or extending rollback segment, TADDR points to this recursive transaction. When your query happens to select at the same time, it sees statistics for the small recursive transaction, not your big one. You should have been joining v$transaction.ses_addr with v$session.saddr instead of v$session.taddr with v$transaction.addr. (note there is a column recursive in v$transaction which allows to filter recursive transactions out) I tested it on 9.2.0.3 with AUM, but it should be the same in earlier versions. Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 4:04 PM 8.1.7.4 No AUM. Either it was something goofy in the database, or something clobbered some bytes as they were going from the host machine to my telnet session ... which afaik would also be wierd since TCP/IP is supposed to guarantee delivery. The terminal session is on a Windows box ... maybe that's it! -Original Message- Which version are you on? Just wondering if it might have something to do with some bug in automatic undo management? Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, September 19, 2003 6:09 PM No question here. Just something weird. This is a long-running insert with NO NONE ZERO ZIP ZILCH NADA commit. It makes me wonder if something weird is going on, or if I am overlooking something in the query. SQL select a.username,sum(b.used_ublk) x from v$session a, v$transaction b where a.taddr=b.addr group by a.username; USERNAMEX -- -- SYSTEM418 1 row selected. SQL / USERNAMEX -- -- SYSTEM893 1 row selected. SQL / USERNAMEX -- -- SYSTEM 2 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3181 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3204 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be
Re: SGA Max size
In 9202, DB_BUBBER_CACHE, SHARED_POOL, LARGE_POOL and JAVA_POOL can be dynamically altered. But in 901, LARGE_POOL and JAVA POOL are static. If MAX SGA is less than 128MB then Oracle will use 4MB granule size to allocate/deallocate memory. For SGA greater than 128M, Oracle granule size is 16MB. Btw, in Windows the granule size is 8M if SGA_MAX_SIZE is set larger than 128M. Note that there's more stuff in SGA than above mentioned areas. Fixed SGA for example etc.. Tanel. -Original Message- To: Multiple recipients of list ORACLE-L Sent: 9/20/03 11:19 AM My understanding of SGA is SGA = x + y + z where x = (dbblksize*db_blk_buf OR db_cache_Size if 9i) y=shared_pool z=java pool, log_buffer If 9i oracle introduced SGA_MAX_SIZE; the sum of x+y+z can be SGA_MAX_SIZE; if so, which part of x/y/z expands when need arises. Thanks Quriyat ___ No banners. No pop-ups. No kidding. Introducing My Way - http://www.myway.com -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: quriyat INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). DISCLAIMER: This message is intended for the sole use of the individual to whom it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not the addressee you are hereby notified that you may not use, copy, disclose, or distribute to anyone the message or any information contained in the message. If you have received this message in error, please immediately advise the sender by reply email and delete this message. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: [EMAIL PROTECTED] INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
Hi Peter, Tanel and Jared, Peter: I meant a public unwrap process not the internal mechanisms in the PL/SQL interpreter / VM. Tanel: I would be more worried about Oracle coming after you in the legally sense if you did reverse engineer the wrap process!! Jared: Are you sure that's how it works? do you have inside knowledge? - if it is this way, is it compiled P-Code or the intermediate DIANA representation? - if it were DIANA or p-code then Peter is wrong above as i would assume that instead of needing an un-wrapper that the VM / interpreter just loads p-code rather than calling the compiler first - if it is DIANA representation then that would mean loading somewhere in the middle of the normal process - or would it? - Is normal (non wrapped) pl/sql that is loaded into the cache held as p-code or DIANA - (or both?). I understood that the wrap process encoded or rather obfuscated the PL/SQL not encrypted it - I am not sure storing it as P-Code or diana would be secure as it should then be possible to extract enough structural program info from the database with the diana packages? or from the tables where the diana - or p-code is held. Anyway's Peter is right in some sense as I heard that some Russian guy is supposed to have reverse engineered the wrap process and un-encoded / decrypted all of the builtin packages and posted the code somewhere on the net - A guy from a security company in the states told me this some months ago but i haven't seen any discussion of it to confirm it. kind regards Pete -- Pete Finnigan email:[EMAIL PROTECTED] Web site: http://www.petefinnigan.com - Oracle security audit specialists Book:Oracle security step-by-step Guide - see http://store.sans.org for details. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Pete Finnigan INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
Tanel Poder wrote: Perhaps you're not aware of the way executables compiled on your Solaris and Windows platforms. In detail, not. In general, yes. Ok, I checked, you're correct, wrap isn't only this 40kB executable, uses orancrypt9.dll (100kB) in Windows, this might be the one where encryption is done... The word 'ecryption' is so amazing and enigmatic, probably that's why so many people are 'poisoned'. It shouldn't be that hard to reverse engineer it. It's an extremely commendable plan... (a touch of irony here) :) I've dealt with disassembling before, back in old dos times (disassembling 4kB graphical intros and few viruses :). I don't think this is a hard job to JFYI, people who made 4kb demos do share their code and ideas, in case one's really interested to get into this. do, it's just time consuming - it gets hard when the authors have planted debugger traps and various other tricks into the code that make the crackers life hard (or should I say interesting :) :) I just think you have/had too much time and nothing serious and important to do. -- Vladimir Begun The statements and opinions expressed here are my own and do not necessarily represent those of Oracle Corporation. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Vladimir Begun INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
minus vs. where not exists, vs. where not in
Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if the subquery is relatively close in cost to the outer query. Yes I know you cant make broad generalizations, but there has to be some 'narrow' generalizations you can make. Such as certain cases, etc... what have you seen? There seems to be very little work in this area in the literature. Does 'where not exist' need more or less sort_area_size space than minus?
RE: weird
Yeah. What he said. And you make the rest of us look bad. For the record, I replaced my ailing 27-inch color TV with a new 32-inch and started a new game of Final Fantasy VII. No stinkin' Oracle for me this weekend. -Original Message- Tanel, This is above and beyond the call of duty. Now you need to step away from the computer... Jared On Sun, 2003-09-21 at 08:29, Tanel Poder wrote: Hi! After spending half of Saturday and Sunday digging around x$ktcxb and other interesting views I actually realized that the answer is way simpler :) TADDR column in v$session points to *current* transaction state object, that means if recursive transaction is needed for wrapping or extending rollback segment, TADDR points to this recursive transaction. When your query happens to select at the same time, it sees statistics for the small recursive transaction, not your big one. You should have been joining v$transaction.ses_addr with v$session.saddr instead of v$session.taddr with v$transaction.addr. (note there is a column recursive in v$transaction which allows to filter recursive transactions out) I tested it on 9.2.0.3 with AUM, but it should be the same in earlier versions. Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 4:04 PM 8.1.7.4 No AUM. Either it was something goofy in the database, or something clobbered some bytes as they were going from the host machine to my telnet session ... which afaik would also be wierd since TCP/IP is supposed to guarantee delivery. The terminal session is on a Windows box ... maybe that's it! -Original Message- Which version are you on? Just wondering if it might have something to do with some bug in automatic undo management? Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, September 19, 2003 6:09 PM No question here. Just something weird. This is a long-running insert with NO NONE ZERO ZIP ZILCH NADA commit. It makes me wonder if something weird is going on, or if I am overlooking something in the query. SQL select a.username,sum(b.used_ublk) x from v$session a, v$transaction b where a.taddr=b.addr group by a.username; USERNAMEX -- -- SYSTEM418 1 row selected. SQL / USERNAMEX -- -- SYSTEM893 1 row selected. SQL / USERNAMEX -- -- SYSTEM 2 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3181 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3204 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat
Re: minus vs. where not exists, vs. where not in
First off, the three are not equivalent, not substitutes for each other. Well not in and minus would be, but they are different from not exists. not in/minus and not exists can return different results. See http://asktom.oracle.com/pls/ask/f?p=4950:8:F4950_P8_DISPLAYID:442029737684http://asktom.oracle.com/pls/ask/f?p=4950:8:F4950_P8_DISPLAYID:442029737684 for examples and explanation. I have not done any performance comparisons but I personally routinely use minus and I am quite happy with it, especially across a db link. At 09:59 AM 9/21/2003 -0800, you wrote: Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if the subquery is relatively close in cost to the outer query. Yes I know you cant make broad generalizations, but there has to be some 'narrow' generalizations you can make. Such as certain cases, etc... what have you seen? There seems to be very little work in this area in the literature. Does 'where not exist' need more or less sort_area_size space than minus? Wolfgang Breitling Oracle7, 8, 8i, 9i OCP DBA Centrex Consulting Corporation http://www.centrexcc.com -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Wolfgang Breitling INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: Copying statistics : used a lot ????
What a timely thread! We've been discussing this very issue back and forth for some time now in our organization. We are getting a new application, which presumably will run on something like lower-end Superdome 16-way, ~30GB RAM type of box. The question is whether 4-way, 4GB RAM Rp5470 (entry-level) will do as a test server or it should be a close mirror of the production one? The argument of those against a similar to a Prod box is simple: Can't afford another one for this project. Don't you know how to use dbms_stats to convince CBO it's on Superdome with 16 CPUs and millions of rows of data and not on a 4-way, couple of thousands in row sources? So if we can't afford what Raj describes - is a smaller server a viable solution for a test box? Or we have to convince damanagement that their can't afford is going to cost them more in the long run? (easlier said than done) --- As Cris mentioned I've read Tom's take on this, but it only confused me futher. Tom states: Some people adopt the strategy of importing the prod statistics ... and think they can get optimizer to generate the plans that will be used in prod and test using that data ... That approach will work only if you can read a query plan and be 100% confident that the plan is good and will give subsecond response times ... I don't think I can make such a judgment call... I don't follow. Does this imply that with importing stats we can't get 100% identical CBO decisions/executions plans in a DB on a smaller machine? Or is it that we have to be 100% confident that we replicated all the stats from Prod and it is not a simple task? Or something else? ... Most people are striving to get query plans that use indexes all of the time, without realizing that as you scale up, indexes may not be the best solution .. This part I understood even less. After reading Cary's excellent paper on scalability I thought that O(n) type of scalability of FTS is worse than say O(log2, n) of IRS? Wouldn't it be correct to say than, that if today on a thousand row tables I get index access path delivering better response time that table scan, I can expect this to stay the same (or better) when my data gets to a million rows range? Is it the scalability of NL vs HJ Tom is taking about? Bitmap/Domain indexes? Or is it a general statement? ... This is not to say that ... import statistics is not very useful. Quite the contrary - I've seen people use (with great success) the ability to import/export statistics, but ***not to tune in test***. Instead they take the results of statistics gathering done in test and import into production! Quite the reverse of what most people initially consider using dbms_stats for... The last remark certainly applies to me. With all due respect to Tom, I got only more confused ater reading the above. Can somebody enlighten me? TIA, Boris Dali. --- Jamadagni, Rajendra [EMAIL PROTECTED] wrote: Our production and test systems are same ... test lags 24 hours behind production that's all. But I have successfully used dbms_stats to copy over stats from production to test on a table by table basis to verify explain plans. My opinion WAD - Works as designed ... remember to take a backup of existing stats on test in a separate table so you can reload them when needed quickly. My experience is on 9ir2 only for this feature. Raj Rajendra dot Jamadagni at nospamespn dot com All Views expressed in this email are strictly personal. QOTD: Any clod can have facts, having an opinion is an art ! -Original Message- Sent: Friday, September 12, 2003 10:29 AM To: Multiple recipients of list ORACLE-L Well no, I have bought the Expert one on one book. I check on his web site and I found one reference where he addresses the use of changing the stats. Usually you can find what is its opinion just by the tone, but this time I was not able to see if he's against or not on this. Can you share more of what's in the book ? Stephane Paquette Administrateur de bases de donnees Database Administrator Standard Life www.standardlife.ca Tel. (514) 499-7999 7470 and (514) 925-7187 [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] -Original Message- Grabowy, Chris Sent: Thursday, September 11, 2003 5:44 PM To: Multiple recipients of list ORACLE-L By chance, do you have Tom Kyte's latest book? Effective Oracle by Design?? He states his opinion on this approach on page 30, section entitled Test Against Representative Data. -Original Message- Stephane Paquette Sent: Thursday, September 11, 2003 4:38 PM To: Multiple recipients of list ORACLE-L HI, I was wandering if a lot of people are copying statistics using dbms_stats from production to test environment to see what will be the access plan. If not used, why ? no time to look at it, bugged, not usefull ,... ?
RE: Copying statistics : used a lot ????
What a timely thread! We've been discussing this very issue back and forth for some time now in our organization. We are getting a new application, which presumably will run on something like lower-end Superdome 16-way, ~30GB RAM type of box. The question is whether 4-way, 4GB RAM Rp5470 (entry-level) will do as a test server or it should be a close mirror of the production one? The argument of those against a similar to a Prod box is simple: Can't afford another one for this project. Don't you know how to use dbms_stats to convince CBO it's on Superdome with 16 CPUs and millions of rows of data and not on a 4-way, couple of thousands in row sources? So if we can't afford what Raj describes - is a smaller server a viable solution for a test box? Or we have to convince damanagement that their can't afford is going to cost them more in the long run? (easlier said than done) --- As Cris mentioned I've read Tom's take on this, but it only confused me futher. Tom states: Some people adopt the strategy of importing the prod statistics ... and think they can get optimizer to generate the plans that will be used in prod and test using that data ... That approach will work only if you can read a query plan and be 100% confident that the plan is good and will give subsecond response times ... I don't think I can make such a judgment call... I don't follow. Does this imply that with importing stats we can't get 100% identical CBO decisions/executions plans in a DB on a smaller machine? Or is it that we have to be 100% confident that we replicated all the stats from Prod and it is not a simple task? Or something else? ... Most people are striving to get query plans that use indexes all of the time, without realizing that as you scale up, indexes may not be the best solution ... This part I understood even less. After reading Cary's excellent paper on scalability I thought that O(n) type of scalability of FTS is worse than say O(log2, n) of IRS? Wouldn't it be correct to say than, that if today on a thousand row tables I get index access path delivering better response time that table scan, I can expect this to stay the same (or better) when my data gets to a million rows range? Is it the scalability of NL vs HJ Tom is taking about? Bitmap/Domain indexes? Or is it a general statement? ... This is not to say that ... import statistics is not very useful. Quite the contrary - I've seen people use (with great success) the ability to import/export statistics, but ***not to tune in test***. Instead they take the results of statistics gathering done in test and import into production! Quite the reverse of what most people initially consider using dbms_stats for... The last remark certainly applies to me. With all due respect to Tom, I got only more confused ater reading the above. Can somebody enlighten me? TIA, Boris Dali. --- Jamadagni, Rajendra [EMAIL PROTECTED] wrote: Our production and test systems are same ... test lags 24 hours behind production that's all. But I have successfully used dbms_stats to copy over stats from production to test on a table by table basis to verify explain plans. My opinion WAD - Works as designed ... remember to take a backup of existing stats on test in a separate table so you can reload them when needed quickly. My experience is on 9ir2 only for this feature. Raj Rajendra dot Jamadagni at nospamespn dot com All Views expressed in this email are strictly personal. QOTD: Any clod can have facts, having an opinion is an art ! -Original Message- Sent: Friday, September 12, 2003 10:29 AM To: Multiple recipients of list ORACLE-L Well no, I have bought the Expert one on one book. I check on his web site and I found one reference where he addresses the use of changing the stats. Usually you can find what is its opinion just by the tone, but this time I was not able to see if he's against or not on this. Can you share more of what's in the book ? Stephane Paquette Administrateur de bases de donnees Database Administrator Standard Life www.standardlife.ca Tel. (514) 499-7999 7470 and (514) 925-7187 [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] -Original Message- Grabowy, Chris Sent: Thursday, September 11, 2003 5:44 PM To: Multiple recipients of list ORACLE-L By chance, do you have Tom Kyte's latest book? Effective Oracle by Design?? He states his opinion on this approach on page 30, section entitled Test Against Representative Data. -Original Message- Stephane Paquette Sent: Thursday, September 11, 2003 4:38 PM To: Multiple recipients of list ORACLE-L HI, I was wandering if a lot of people are copying statistics using dbms_stats from production to test environment to see what will be the access plan. If not used, why ? no time to look at it, bugged, not usefull ,... ?
Re: minus vs. where not exists, vs. where not in
if you handle for nulls with an 'nvl' then 'not exists' appears to return the same answer as not in. or am I wrong? - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 3:24 PM First off, the three are not equivalent, not substitutes for each other. Well not in and minus would be, but they are different from not exists. not in/minus and not exists can return different results. See http://asktom.oracle.com/pls/ask/f?p=4950:8:F4950_P8_DISPLAYID:44202973 7684http://asktom.oracle.com/pls/ask/f?p=4950:8:F4950_P8_DISPLAYID:4420 29737684 for examples and explanation. I have not done any performance comparisons but I personally routinely use minus and I am quite happy with it, especially across a db link. At 09:59 AM 9/21/2003 -0800, you wrote: Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if the subquery is relatively close in cost to the outer query. Yes I know you cant make broad generalizations, but there has to be some 'narrow' generalizations you can make. Such as certain cases, etc... what have you seen? There seems to be very little work in this area in the literature. Does 'where not exist' need more or less sort_area_size space than minus? Wolfgang Breitling Oracle7, 8, 8i, 9i OCP DBA Centrex Consulting Corporation http://www.centrexcc.com -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Wolfgang Breitling INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Ryan INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: Copying statistics : used a lot ????
wait a second. the CBO takes into consideration your system statistics when you analyze? Is that new in 9i? I thought the export stats and import stats were used if you wanted a 'smaller subset' of data. so you mimic the data stats? - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 3:34 PM What a timely thread! We've been discussing this very issue back and forth for some time now in our organization. We are getting a new application, which presumably will run on something like lower-end Superdome 16-way, ~30GB RAM type of box. The question is whether 4-way, 4GB RAM Rp5470 (entry-level) will do as a test server or it should be a close mirror of the production one? The argument of those against a similar to a Prod box is simple: Can't afford another one for this project. Don't you know how to use dbms_stats to convince CBO it's on Superdome with 16 CPUs and millions of rows of data and not on a 4-way, couple of thousands in row sources? So if we can't afford what Raj describes - is a smaller server a viable solution for a test box? Or we have to convince damanagement that their can't afford is going to cost them more in the long run? (easlier said than done) --- As Cris mentioned I've read Tom's take on this, but it only confused me futher. Tom states: Some people adopt the strategy of importing the prod statistics ... and think they can get optimizer to generate the plans that will be used in prod and test using that data ... That approach will work only if you can read a query plan and be 100% confident that the plan is good and will give subsecond response times ... I don't think I can make such a judgment call... I don't follow. Does this imply that with importing stats we can't get 100% identical CBO decisions/executions plans in a DB on a smaller machine? Or is it that we have to be 100% confident that we replicated all the stats from Prod and it is not a simple task? Or something else? ... Most people are striving to get query plans that use indexes all of the time, without realizing that as you scale up, indexes may not be the best solution .. This part I understood even less. After reading Cary's excellent paper on scalability I thought that O(n) type of scalability of FTS is worse than say O(log2, n) of IRS? Wouldn't it be correct to say than, that if today on a thousand row tables I get index access path delivering better response time that table scan, I can expect this to stay the same (or better) when my data gets to a million rows range? Is it the scalability of NL vs HJ Tom is taking about? Bitmap/Domain indexes? Or is it a general statement? ... This is not to say that ... import statistics is not very useful. Quite the contrary - I've seen people use (with great success) the ability to import/export statistics, but ***not to tune in test***. Instead they take the results of statistics gathering done in test and import into production! Quite the reverse of what most people initially consider using dbms_stats for... The last remark certainly applies to me. With all due respect to Tom, I got only more confused ater reading the above. Can somebody enlighten me? TIA, Boris Dali. --- Jamadagni, Rajendra [EMAIL PROTECTED] wrote: Our production and test systems are same ... test lags 24 hours behind production that's all. But I have successfully used dbms_stats to copy over stats from production to test on a table by table basis to verify explain plans. My opinion WAD - Works as designed ... remember to take a backup of existing stats on test in a separate table so you can reload them when needed quickly. My experience is on 9ir2 only for this feature. Raj -- -- Rajendra dot Jamadagni at nospamespn dot com All Views expressed in this email are strictly personal. QOTD: Any clod can have facts, having an opinion is an art ! -Original Message- Sent: Friday, September 12, 2003 10:29 AM To: Multiple recipients of list ORACLE-L Well no, I have bought the Expert one on one book. I check on his web site and I found one reference where he addresses the use of changing the stats. Usually you can find what is its opinion just by the tone, but this time I was not able to see if he's against or not on this. Can you share more of what's in the book ? Stephane Paquette Administrateur de bases de donnees Database Administrator Standard Life www.standardlife.ca Tel. (514) 499-7999 7470 and (514) 925-7187 [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] -Original Message- Grabowy, Chris Sent: Thursday, September 11, 2003 5:44 PM To: Multiple recipients of list ORACLE-L By chance, do you have Tom Kyte's latest book? Effective Oracle by Design?? He
Re: wrapping packages
:) I just think you have/had too much time and nothing serious and important to do. That was the case, back at highschool days... Tanel. -- Vladimir Begun The statements and opinions expressed here are my own and do not necessarily represent those of Oracle Corporation. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Vladimir Begun INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: minus vs. where not exists, vs. where not in
On 2003.09.21 15:24, Wolfgang Breitling wrote: First off, the three are not equivalent, not substitutes for each other. Well not in and minus would be, but they are different from not exists. not in/minus and not exists can return different results. See http://asktom.oracle.com/pls/ask/f?p=4950:8:F4950_P8_DISPLAYID:442029737684http://asktom.oracle.com/pls/ask/f?p=4950:8:F4950_P8_DISPLAYID:442029737684 for examples and explanation. I have not done any performance comparisons but I personally routinely use minus and I am quite happy with it, especially across a db link. Of course you do, because oracle brings the results of the whole remote query over db link into the temporary tablespace of the local database. Query like select ename,job,dname from emp e, [EMAIL PROTECTED] d where e.deptno=d.deptno will bring the whole dept table over the database link into the temporary tablespace and perform join. The not exist condition may be faster if indexes are involved and if nested loops will give better results then sort/ merge, but those cases have to be carefully optimized and measured. Now, a slight digression: exactly because of the database having tendency to bring a ton of information over the database link, I frequently try to access remote views to bring over just a few necessary records. -- Mladen Gogala Oracle DBA -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Mladen Gogala INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: weird
Sorry, I can't help it. But anyway, I will take your advice and keep away from Oracle for a while. There are few DB2 materials I want to go through... (just joking ;) Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 10:19 PM Yeah. What he said. And you make the rest of us look bad. For the record, I replaced my ailing 27-inch color TV with a new 32-inch and started a new game of Final Fantasy VII. No stinkin' Oracle for me this weekend. -Original Message- Tanel, This is above and beyond the call of duty. Now you need to step away from the computer... Jared On Sun, 2003-09-21 at 08:29, Tanel Poder wrote: Hi! After spending half of Saturday and Sunday digging around x$ktcxb and other interesting views I actually realized that the answer is way simpler :) TADDR column in v$session points to *current* transaction state object, that means if recursive transaction is needed for wrapping or extending rollback segment, TADDR points to this recursive transaction. When your query happens to select at the same time, it sees statistics for the small recursive transaction, not your big one. You should have been joining v$transaction.ses_addr with v$session.saddr instead of v$session.taddr with v$transaction.addr. (note there is a column recursive in v$transaction which allows to filter recursive transactions out) I tested it on 9.2.0.3 with AUM, but it should be the same in earlier versions. Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 4:04 PM 8.1.7.4 No AUM. Either it was something goofy in the database, or something clobbered some bytes as they were going from the host machine to my telnet session ... which afaik would also be wierd since TCP/IP is supposed to guarantee delivery. The terminal session is on a Windows box ... maybe that's it! -Original Message- Which version are you on? Just wondering if it might have something to do with some bug in automatic undo management? Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Friday, September 19, 2003 6:09 PM No question here. Just something weird. This is a long-running insert with NO NONE ZERO ZIP ZILCH NADA commit. It makes me wonder if something weird is going on, or if I am overlooking something in the query. SQL select a.username,sum(b.used_ublk) x from v$session a, v$transaction b where a.taddr=b.addr group by a.username; USERNAMEX -- -- SYSTEM418 1 row selected. SQL / USERNAMEX -- -- SYSTEM893 1 row selected. SQL / USERNAMEX -- -- SYSTEM 2 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3181 1 row selected. SQL / USERNAMEX -- -- SYSTEM 3204 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Stephen Lee INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an
Re: wrapping packages
Hi! I think disassembling the code itself wouldn't be that condemnable (is this correct usage of the word?), but if anyone would start disributing the wrapping algorithm or spreading modified Oracle code, that would make Oracle wake up. After all, I do have the right to know, which code is executed on my computer (OTOH, I've not read any agreements too thoroughly, when downloading software). Anyway, I don't have that much money to spend on lawyers than Oracle does, so I won't start spreading the results. And I probably won't have any results anyway, since I don't have that much spare time anymore, like Vladimir pointed out... Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Sunday, September 21, 2003 8:39 PM Hi Peter, Tanel and Jared, Peter: I meant a public unwrap process not the internal mechanisms in the PL/SQL interpreter / VM. Tanel: I would be more worried about Oracle coming after you in the legally sense if you did reverse engineer the wrap process!! Jared: Are you sure that's how it works? do you have inside knowledge? - if it is this way, is it compiled P-Code or the intermediate DIANA representation? - if it were DIANA or p-code then Peter is wrong above as i would assume that instead of needing an un-wrapper that the VM / interpreter just loads p-code rather than calling the compiler first - if it is DIANA representation then that would mean loading somewhere in the middle of the normal process - or would it? - Is normal (non wrapped) pl/sql that is loaded into the cache held as p-code or DIANA - (or both?). I understood that the wrap process encoded or rather obfuscated the PL/SQL not encrypted it - I am not sure storing it as P-Code or diana would be secure as it should then be possible to extract enough structural program info from the database with the diana packages? or from the tables where the diana - or p-code is held. Anyway's Peter is right in some sense as I heard that some Russian guy is supposed to have reverse engineered the wrap process and un-encoded / decrypted all of the builtin packages and posted the code somewhere on the net - A guy from a security company in the states told me this some months ago but i haven't seen any discussion of it to confirm it. kind regards Pete -- Pete Finnigan email:[EMAIL PROTECTED] Web site: http://www.petefinnigan.com - Oracle security audit specialists Book:Oracle security step-by-step Guide - see http://store.sans.org for details. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Pete Finnigan INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: log miner views contain object IDs and HEX values
Steve,
Did you have set serveroutput on?
I have found you need this before you'll get error messages from Logminer.
I found this in Metalink note Note:69606.1
HTH in the future,
Bruce Reardon
-Original Message-
Sent: Saturday, 20 September 2003 9:40 AM
OK seems I must have fat fingered something when I typed in my dictionary
location in start_logmnr. Odd though I saw no errors.
-Original Message-
Steve McClure
Sent: Friday, September 19, 2003 1:25 PM
OK my first usage of log miner seemed to go exactly by the book until I went
to look at the data in v$logmnr_contents. The sql_redo column was filled
with strings like the following.
update UNKNOWN.Objn:3167 set Col[43] = HEXTORAW('7867090f010101') ...
I created the dictionary file, and didn't get any errors when I referenced
it in the start_logmnr procedure. I understand that the dictionary file is
required to get my actual object names and column names and such. I am also
pretty sure it was created correctly, so it shouldn't be the root of my
problem.
As I am typing this, thinking it over, I realize that I was logged in as my
own user connected as sysdba. Maybe I should redo this connected as ths sys
user. I created the dictionary file as the sys user. I am gonna have to
try that after lunch. While I do though, and since I have already typed
this much, I am gonna toss this up to my fellow listers. The only
documentation I have been using is the 8i Administrators Guide, and one of
its examples references a column that isn't in the v$logmnr_contents view.
Perhaps I need a different reference.
Thanks,
Steve
Author: Steve McClure
INET: [EMAIL PROTECTED]
--
Please see the official ORACLE-L FAQ: http://www.orafaq.net
--
Author: Reardon, Bruce (CALBBAY)
INET: [EMAIL PROTECTED]
Fat City Network Services-- 858-538-5051 http://www.fatcity.com
San Diego, California-- Mailing list and web hosting services
-
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from). You may
also send the HELP command for other information (like subscribing).
Re: minus vs. where not exists, vs. where not in
I don't have any scientific proof but I imagine the sort_area_size could be different between not exists and minus. When doing minus the entire select string would have to be compared for equality, which could be very large depending on your string. Minus is probably an easier way (at least from a coding perspective) of comparing every column in a table, or incorporating complexity in the select statement. Having said that. Each statement has it's pros and cons and I guess it depends on the volume of the various queries, etc. For example, if the have table A with 5 rows and table B with 100 rows executing select a.field from a where not exists (select 1 from b where b.field = a.field) will result in 5 quick searches (assuming b.field is indexed) whereas select a.field from a where a.field not in (select b.field from b) will result in a full index read of b, as would select a.field from a minus select b.field from b. Obviously different queries and different volumes will tip the scales in different directions. Regards, Mark. Ryan [EMAIL PROTECTED]To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] cc: Sent by: Subject: minus vs. where not exists, vs. where not in [EMAIL PROTECTED] .com 22/09/2003 03:59 Please respond to ORACLE-L Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if the subquery is relatively close in cost to the outer query. Yes I know you cant make broad generalizations, but there has to be some 'narrow' generalizations you can make. Such as certain cases, etc... what have you seen? There seems to be very little work in this area in the literature. Does 'where not exist' need more or less sort_area_size space than minus? Privileged/Confidential information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such case, you should destroy this message and kindly notify the sender by reply e-mail or by telephone on (61 3) 9612-6999. Please advise immediately if you or your employer does not consent to Internet e-mail for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of Transurban City Link Ltd shall be understood as neither given nor endorsed by it. Privileged/Confidential information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such a case, you should destroy this message and kindly notify the sender by reply e-mail or by telephone on (03) 9612-6999 or (61) 3 9612-6999. Please advise immediately if you or your employer does not consent to Internet e-mail for
Re: Anyone have a copy of DUL ??
Bernard of Oracle Holland made DUL (Direct UnLoader) several years ago. I took the very last internals class conducted by Oracle Support EMEA Vice President Andre Bakker (the only VP to conduct internals classes, I think. I also think he quit Support in disgust some time later :-) ), and we talked about a severe case we had in Denmark at that time. Basically, a company that made technical specs (including drawings) for some very advanced, high-speed transportation things, had not taken a backup of their system tablespace for 18 months. Then the system01.dbf file did its own thing, and their database didn't really feel good. Andre told me that he had a guy who was working on a tool that might be able to help us out. But it was in beta, etc., etc. So I went back to Denmark, called Bernhard, and we agreed that he would fix the bugs as we encountered them. I sent one of my guys - Christian Fabricius - online, and he was gone for three days, but got all the technical drawings out of the datafiles (Bernhard had to fix two or three things as we went along - a tribute to his coding skills). All that time, practially, Bernhard was online. Rock'n'roll. We were very proud. First time in history. Blah blah blah. When my manager went to a meeting with the customer a week later we were all expecting joy and happiness and perhaps some gratefullness from the customer. But no. He was furious. Why hadn't we told him that it was neccessary to take a backup of the system tablespace? Where in the documentation did it clearly state that that was required? It was all our fault. And we didn't even charge them more than normal hourly rates. I never tried the other suggestion Andre had (and which he had used many times himself): Create a dummy database that has the same datafiles as the problem database. Then take the file headers from the dummy database and patch on top of the real database. Then you can start up, since the information in the file headers match. Andre was one cool guy. He's enjoying early retirement, he claims. Mogens Rachel Carmichael wrote: Kevin Loney tells the story of making a call to the data center from the CIO's office and asking them to make a copy of the backup tapes and leave them at reception. since the call came from the CIO's office, they made the copy --- Pete Finnigan [EMAIL PROTECTED] wrote: Hi Peter Glad to hear that there are controls in Oracle for use of DUL, I was thinking of a case where i heard that one guy rang up the backup storage company for a large company and requested a set of backup tapes be left at reception at the company and he just walked in off the street and took them. Mitnik tells similar stories in his book. Thanks for the internal Oracle insight Peter, kind regards Pete In article [EMAIL PROTECTED], Peter Gram [EMAIL PROTECTED] writes Hi Pete I have used Dul many times at customer sites when I was employed by Oracle Denmark. Every time the customer management had to verify by phone and fax that they understood the full impact of using Dul. Oracle have disclaimer that explains the problems with missing transaction consistency of the data saved by Dul and the security issues. The customer has to sign and fax the disclaimer back to Oracle before we came on site .-) After I left Oracle several people ask me if would write a Dul and I declined. I'm of the opinion that Dul should stay behind the Oracle firewall. /peter Pete Finnigan wrote: Hi Mark I agree with you Mark, even if its supplied by Oracle technicians - it is as you say possible to by-pass security completely. Does anyone in Oracle check that the field support personnel dispatched to a site ( in urgency ) are dumping data for the owner of it? - I covered the issue of DUL with regards to security is the SANS Oracle security step-by-step book - action 6.5.1 kind regards Pete In article [EMAIL PROTECTED], Mark Leith [EMAIL PROTECTED] writes One problem I see with giving this away free is that you will be supplying a tool that allows you to extract data from the database, bypassing all inbuilt security. A BIG no no. I suppose that also applies to this kind of tool even under a paid license structure. -- Pete Finnigan email:[EMAIL PROTECTED] Web site: http://www.petefinnigan.com - Oracle security audit specialists Book:Oracle security step-by-step Guide - see http://store.sans.org for details. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Pete Finnigan INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED]
Re: minus vs. where not exists, vs. where not in
I'm not sure that there is a good answer to that question. The question sounds like the dilemma: who's stronger, Batman or Superman? Unfortunately, superheroes do not exist, so we cannot have a real life comparison. It is exactly the same with not exists vs. minus. Comparisons make sense only within a real life application in a real configuration, that is why both mechanisms are provided. Sort_area_size does influence performance, unless memory is slow, system is swapping or something else. In the world of superheroes, my favorite is Alice and her fist of death. On 2003.09.21 19:54, Mark Richard wrote: I don't have any scientific proof but I imagine the sort_area_size could be different between not exists and minus. When doing minus the entire select string would have to be compared for equality, which could be very large depending on your string. Minus is probably an easier way (at least from a coding perspective) of comparing every column in a table, or incorporating complexity in the select statement. Having said that. Each statement has it's pros and cons and I guess it depends on the volume of the various queries, etc. For example, if the have table A with 5 rows and table B with 100 rows executing select a.field from a where not exists (select 1 from b where b.field = a.field) will result in 5 quick searches (assuming b.field is indexed) whereas select a.field from a where a.field not in (select b.field from b) will result in a full index read of b, as would select a.field from a minus select b.field from b. Obviously different queries and different volumes will tip the scales in different directions. Regards, Mark. Ryan [EMAIL PROTECTED]To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] cc: Sent by: Subject: minus vs. where not exists, vs. where not in [EMAIL PROTECTED] .com 22/09/2003 03:59 Please respond to ORACLE-L Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if the subquery is relatively close in cost to the outer query. Yes I know you cant make broad generalizations, but there has to be some 'narrow' generalizations you can make. Such as certain cases, etc... what have you seen? There seems to be very little work in this area in the literature. Does 'where not exist' need more or less sort_area_size space than minus? Privileged/Confidential information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such case, you should destroy this message and kindly notify the sender by reply e-mail or by telephone on (61 3) 9612-6999. Please advise immediately if you or your employer does not consent to Internet e-mail for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of Transurban City Link Ltd shall be understood as neither given nor endorsed by it. Privileged/Confidential information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such a case, you should destroy this message and kindly notify the sender by reply e-mail or by telephone on (03) 9612-6999 or (61) 3 9612-6999. Please advise immediately if you or your employer does not consent to Internet e-mail for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of Transurban Infrastructure Developments Limited and CityLink Melbourne Limited shall be understood as neither given nor endorsed by them. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Mark Richard INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing
RE: wrapping packages
Jared, That is also my understanding of what wrap does - after all if the wrapped PL/SQL code is encrypted why do string literals appear within the wrapped output. And where do you specify the encryption key - you don't because wrap does not encrypt. I'm not surprised that people think the wrap command encrypts the code seeing as authors likes Couchman and Marisetti use this language within OCP Oracle9i Database Fundamentals I Exam Guide page 96 and it is also stated in the PL/SQL User's Guide and Reference. Cheers, Craig. -Original Message- Sent: Monday, 22 September 2003 2:15 AM To: Multiple recipients of list ORACLE-L IIRC, 'wrap' does not actually encrypt the code. Rather, it simply does a precompile on it and then stores the pcode in the database. Jared On Sat, 2003-09-20 at 14:29, Peter Gram wrote: Hi Pete I must point out that there must be a unwrap, since the Oracle database can run the wrapped pl/sql code :-) It is based on trust in Oracle cooperation / development. Some times it would make since to write the code in c/c++ since it harder to revers. Pete Finnigan wrote: Hi Very true, but if there was the wrap process wouldn't be much use as anyone could un wrap your code. But you are right the main reason to be cautious is to not delete your source code locally. kind regards Pete In article [EMAIL PROTECTED], bhabani s pradhan [EMAIL PROTECTED] writes one caution: there is no unwrap cmd/exe Regards -- Peter Gram, Miracle A/S Phone : +45 2527 7107, Fax : +45 4466 8856, Home +45 3874 5696 mail : [EMAIL PROTECTED] - http://MiracleAS.dk Upcoming events: DatabaseForum 2003, Lalandia 2-4 October Visit http://miracleas.dk/events/DBF2003/invitation.html Miracle Master Class with Tom Kyte, 12-14 January 2004 -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Peter Gram INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Jared Still INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Craig Munday INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
sql trace tuning articles
did a google search and couldnt find anything worth reading. other than the ones on hotsos any other good ones? namely ones on traces other than 10053 and 10046? Ive seen a few others mentioned but no details.
Re: sql trace tuning articles
Did you check out www.hotsos.com? Cary Millsap and Jeff Holt have a few (as well as a great book Optimizing Oracle Performance) Thanks/Richard --- Ryan [EMAIL PROTECTED] wrote: did a google search and couldnt find anything worth reading. other than the ones on hotsos any other good ones? namely ones on traces other than 10053 and 10046? Ive seen a few others mentioned but no details. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Richard Stroupe INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
Tanel Poder wrote: :) I just think you have/had too much time and nothing serious and important to do. That was the case, back at highschool days... I think you're still there... at least according to your posts. :) -- Vladimir Begun The statements and opinions expressed here are my own and do not necessarily represent those of Oracle Corporation. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Vladimir Begun INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
Anyway's Peter is right in some sense as I heard that some Russian guy Those Russians... :) They can do a lot. -- Vladimir Begun The statements and opinions expressed here are my own and do not necessarily represent those of Oracle Corporation. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Vladimir Begun INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: wrapping packages
Tanel Poder wrote: After all, I do have the right to know, which code is executed on my computer (OTOH, I've not read any agreements too thoroughly, when downloading software). I do have the right to know which code is executed on my computer or not execute that code but not hack it to know what's running there. Read agreements :) [and I would not suggest you to discuss illegal things -- 'how to hack' -- it creates wrong impression about you as about an IT person, IMHO. Intelligent IT individual c00l hazker. Believe me, I know what I'm talking about, it's much more cool when you knock the door (in this case [EMAIL PROTECTED]) than someone else knocks your door -- feel the difference (c)] -- Vladimir Begun The statements and opinions expressed here are my own and do not necessarily represent those of Oracle Corporation. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Vladimir Begun INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: minus vs. where not exists, vs. where not in
I think Superman is stroger than Batman -Original Message- Sent: 22 September 2003 08:30 To: Multiple recipients of list ORACLE-L I'm not sure that there is a good answer to that question. The question sounds like the dilemma: who's stronger, Batman or Superman? Unfortunately, superheroes do not exist, so we cannot have a real life comparison. It is exactly the same with not exists vs. minus. Comparisons make sense only within a real life application in a real configuration, that is why both mechanisms are provided. Sort_area_size does influence performance, unless memory is slow, system is swapping or something else. In the world of superheroes, my favorite is Alice and her fist of death. On 2003.09.21 19:54, Mark Richard wrote: I don't have any scientific proof but I imagine the sort_area_size could be different between not exists and minus. When doing minus the entire select string would have to be compared for equality, which could be very large depending on your string. Minus is probably an easier way (at least from a coding perspective) of comparing every column in a table, or incorporating complexity in the select statement. Having said that. Each statement has it's pros and cons and I guess it depends on the volume of the various queries, etc. For example, if the have table A with 5 rows and table B with 100 rows executing select a.field from a where not exists (select 1 from b where b.field = a.field) will result in 5 quick searches (assuming b.field is indexed) whereas select a.field from a where a.field not in (select b.field from b) will result in a full index read of b, as would select a.field from a minus select b.field from b. Obviously different queries and different volumes will tip the scales in different directions. Regards, Mark. Ryan [EMAIL PROTECTED]To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] cc: Sent by: Subject: minus vs. where not exists, vs. where not in [EMAIL PROTECTED] .com 22/09/2003 03:59 Please respond to ORACLE-L Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if the subquery is relatively close in cost to the outer query. Yes I know you cant make broad generalizations, but there has to be some 'narrow' generalizations you can make. Such as certain cases, etc... what have you seen? There seems to be very little work in this area in the literature. Does 'where not exist' need more or less sort_area_size space than minus? Privileged/Confidential information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such case, you should destroy this message and kindly notify the sender by reply e-mail or by telephone on (61 3) 9612-6999. Please advise immediately if you or your employer does not consent to Internet e-mail for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of Transurban City Link Ltd shall be understood as neither given nor endorsed by it. Privileged/Confidential information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such a case, you should destroy this message and kindly notify the sender by reply e-mail or by telephone on (03) 9612-6999 or (61) 3 9612-6999. Please advise immediately if you or your employer does not consent to Internet e-mail for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of Transurban Infrastructure Developments Limited and CityLink Melbourne Limited shall be understood as neither given nor endorsed by them. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Mark Richard INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services
Bug in 9.2.0.4 when order by constant desc
Hi all, Got the info' below from www.dba-village.com and thought of sharing it with u guys. If you try select 'Hello' from dual order by 1 desc; you may run into the bug. It creates a dump in udump and could disconnect your session. Oracle is creating a one-off patch to fix. The bug is for any order by x, where x is a number, desc and the column is a constant. Jp. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Prem Khanna J INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Oracle 9i new features - by Connor McDonald
..a paper presented by Connor at OracleWorld 2003 is here. http://tinyurl.com/o6hp ..thought it may be of interest for many on this list. Jp. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Prem Khanna J INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: minus vs. where not exists, vs. where not in
Yeah, Superman had Super powers. Batman just relied on a bunch of gadgets. Of course my original post was implying the classic it depends. There are some scenario's where a specific approach is faster than the other and there are some scenario's where it doesn't matter because oracle can rewrite the query effectively anyway. What does matter - If in doubt test it out. Sinardy Xing [EMAIL PROTECTED]To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] omsvc.com cc: Sent by: Subject: RE: minus vs. where not exists, vs. where not in [EMAIL PROTECTED] .com 22/09/2003 15:29 Please respond to ORACLE-L I think Superman is stroger than Batman -Original Message- Sent: 22 September 2003 08:30 To: Multiple recipients of list ORACLE-L I'm not sure that there is a good answer to that question. The question sounds like the dilemma: who's stronger, Batman or Superman? Unfortunately, superheroes do not exist, so we cannot have a real life comparison. It is exactly the same with not exists vs. minus. Comparisons make sense only within a real life application in a real configuration, that is why both mechanisms are provided. Sort_area_size does influence performance, unless memory is slow, system is swapping or something else. In the world of superheroes, my favorite is Alice and her fist of death. On 2003.09.21 19:54, Mark Richard wrote: I don't have any scientific proof but I imagine the sort_area_size could be different between not exists and minus. When doing minus the entire select string would have to be compared for equality, which could be very large depending on your string. Minus is probably an easier way (at least from a coding perspective) of comparing every column in a table, or incorporating complexity in the select statement. Having said that. Each statement has it's pros and cons and I guess it depends on the volume of the various queries, etc. For example, if the have table A with 5 rows and table B with 100 rows executing select a.field from a where not exists (select 1 from b where b.field = a.field) will result in 5 quick searches (assuming b.field is indexed) whereas select a.field from a where a.field not in (select b.field from b) will result in a full index read of b, as would select a.field from a minus select b.field from b. Obviously different queries and different volumes will tip the scales in different directions. Regards, Mark. Ryan [EMAIL PROTECTED]To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] cc: Sent by: Subject: minus vs. where not exists, vs. where not in [EMAIL PROTECTED] .com 22/09/2003 03:59 Please respond to ORACLE-L Do any of you have any cases when minus is superior in performance? Ive found 'not in' with a hash_aj to be the best option if the sub-query is significantly less 'costly' then the outer query. I dont mean cost in terms of the Explain Plan, I mean the work Oracle has to do to find a result set. It also depends on a proper hash_area_size I find 'where not exists' to be best if
How do we know that an index need to be rebuilt.
Dear Friends, Can somebody tell me how do we that an index needs to be rebuilt.. Different scenarios / any documents will be helpful. Thanks in advance. Rajuvera ** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ** -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Veeraraju_Mareddi INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: How do we know that an index need to be rebuilt.
Hi Raju, Check the doc. i have attached. HTH. Jp. 22-09-2003 14:59:35, Veeraraju_Mareddi [EMAIL PROTECTED] wrote: Dear Friends, Can somebody tell me how do we that an index needs to be rebuilt.. Different scenarios / any documents will be helpful. Thanks in advance. Rajuvera -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Prem Khanna J INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
RE: oraperf.com is now Veritas
Title: Message All of it? how did you plough through all the detail? Niall -Original Message-From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jamadagni, RajendraSent: 18 September 2003 17:40To: Multiple recipients of list ORACLE-LSubject: RE: oraperf.com is now Veritas I noticed that when I read Anjo's paper at OOW . Raj -Original Message- From: Jesse, Rich [mailto:[EMAIL PROTECTED]] Sent: Thursday, September 18, 2003 11:30 AM To: Multiple recipients of list ORACLE-L Subject: oraperf.com is now Veritas Sort of OT, but it is Oracle information related: Veritas has taken over http://oraperf.com I don't know if this is good, bad, or indifferent, but it's a change that I thought some might find interesting. Or not. Rich Rich Jesse System/Database Administrator [EMAIL PROTECTED] Quad/Tech Inc, Sussex, WI USA -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Jesse, Rich INET: [EMAIL PROTECTED] Fat City Network Services -- 858-538-5051 http://www.fatcity.com San Diego, California -- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: How do we know that an index need to be rebuilt.
Oracle indexes do not need rebuilding on the regular basis, but only as an exceptional event. The warning sign is when queries that utilize nested loops slow down without any apparent reason. The only case when index needs to be rebuilt is the case of a table with frequent delete operations. You should consider rebuilding indexes if and only if the base table satisfies that crieria. How to decide? There are few rules of thumb, none satisfactory and decisive. I usually compare the number of blocks in the index with the number of blocks in the table. If the number of blocks in the index exceeds 50% of those in the table, the index is a candidate for rebuilding. How did I come to 50%? I have no clue. That value popped up at one of my previous jobs and I still use it, when I want to rebuild indexes. One of the most frequent reasons for index rebuilding is a magazine reading boss who has read that indexes need rebuilding. A script which rebuilds based on a criteria like 50% of the number of blocks usually satisfies damagement, especially if accompanied by a 3-page MS-Word document explaining this hocus-pocus in detail. Picture of a B*-tree scanned from N. Wirth's Data Structures+Algorithms = Programs is a mandatory requirement for such documents. On 2003.09.22 01:59, Veeraraju_Mareddi wrote: Dear Friends, Can somebody tell me how do we that an index needs to be rebuilt.. Different scenarios / any documents will be helpful. Thanks in advance. Rajuvera ** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ** -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Veeraraju_Mareddi INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Mladen Gogala Oracle DBA -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Mladen Gogala INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
Re: How do we know that an index need to be rebuilt.
Attachments are stripped from the list messages. On 2003.09.22 02:24, Prem Khanna J wrote: Hi Raju, Check the doc. i have attached. HTH. Jp. 22-09-2003 14:59:35, Veeraraju_Mareddi [EMAIL PROTECTED] wrote: Dear Friends, Can somebody tell me how do we that an index needs to be rebuilt.. Different scenarios / any documents will be helpful. Thanks in advance. Rajuvera -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Prem Khanna J INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Mladen Gogala Oracle DBA -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Mladen Gogala INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
