InnovatorC Bug ?

Posted in 2011

Topics: Storage & Space Management, Server Administration, Triggers, Constraints & Referential Integrity, Licensing & Editions, Migration, Import/Export & Data Conversion, Platform-Specific Issues, Internationalization & Character Sets

Marcus Haarmann — Wed, 16 Mar 2011 16:34:07 — source: IIUG Forums & Mailing Lists

Hi experts, we found an interesting problem while testing with some copied databases from out production system. Following situation: We used an instance of 11.70UC1IE on Linux 32 bit (Ubuntu 10.04) to load some data from production databases (using create table, load from, create index, trigger, foreign keys, done via dbaccess). We encountered the problem, that after the creation of the indexes, oncheck -cI had errors (freshly loaded, index created new). oncheck -cI derpart_eigen:crm_person_phonetic Validating indexes for db:midoco.crm_person_phonetic...Index 111_35 Index fragment partition rootdbs in DBspace rootdbs Index i_crm_person_phonetic01 Index fragment partition rootdbs in DBspace rootdbs Index i_crm_person_phonetic02 Index fragment partition rootdbs in DBspace rootdbs ERROR: Keys in node 0x2 are not in order. ERROR: Index i_crm_person_phonetic02 for db:midoco.crm_person_phonetic is bad. Index i_crm_person_phonetic03 Please Drop and ReCreate Index i_crm_person_phonetic02 for db:midoco.crm_person_phonetic. I assumed a charsset problem and made sure the DB_LOCALE was correct. But data was not corrupt, and also oncheck -cD did not produce any errors. After drop and re-creation of the index oncheck -cI gave the same output ! i_crm_person_phonetic02 is an index over the field forename, which contains uppercase forenames. I also tested with 11.50UC8IE, same results. In my personal 11.50UC5DE (Developer Edition, also Ubuntu 32bit) the problem does not occur. Also in our production environment, the problem does not occur (Workgroup Edition 64bit, 11.10FC3). Thats where the data comes from, but should not make any difference, since we only work on unload data. So I tried to find out if the error is occurring on that machine. But the load with a copy of my Developer edition was OK. I then tried to start up the dbspace from my developer edition with the 11.50UC8IE (before, I did a oncheck -cI which was without errors). After startup of the Innovator Edition (no data changes in between, but internal conversion occurred), oncheck -cI for the same table produced again the above results. Might be a bug in Innovator C ? Any hints ? Marcus Haarmann Geschäftsführer Midoco GmbH Otto-Hahn-Str. 12 40721 Hilden Tel. +49 (2103) 28 74 0 Fax. +49 (2103) 28 74 28 www.midoco.de Member of Pisano Holding GmbH Better Travel Technology www.pisano-holding.com Amtsgericht Düsseldorf - HRB 51420 - USt.ID DE814319276 - Geschäftsführer: Steffen Faradi, Marcus Haarmann, Jörg Hauschild -

↪ replying to Marcus Haarmann

Art Kagel — Wed, 16 Mar 2011 17:10:57 — source: IIUG Forums & Mailing Lists

Sounds like a bug in oncheck. Call IBM and open a case. Art Art S. Kagel Advanced DataTools (www.advancedatatools.com) Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Wed, Mar 16, 2011 at 12:34 PM, Marcus Haarmann <marcus.haarmann@midoco.de > wrote: > Hi experts, > > we found an interesting problem while testing with some copied databases > from out production system. > Following situation: > We used an instance of 11.70UC1IE on Linux 32 bit (Ubuntu 10.04) to load > some data from production databases > (using create table, load from, create index, trigger, foreign keys, > done via dbaccess). > > We encountered the problem, that after the creation of the indexes, > oncheck -cI had errors (freshly loaded, index created new). > oncheck -cI derpart_eigen:crm_person_phonetic > Validating indexes for db:midoco.crm_person_phonetic...> Index 111_35 > > Index fragment partition rootdbs in DBspace rootdbs > Index i_crm_person_phonetic01 > > Index fragment partition rootdbs in DBspace rootdbs > Index i_crm_person_phonetic02 > > Index fragment partition rootdbs in DBspace rootdbs > ERROR: Keys in node 0x2 are not in order. > ERROR: Index i_crm_person_phonetic02 for db:midoco.crm_person_phonetic > is bad. > Index i_crm_person_phonetic03 > Please Drop and ReCreate Index i_crm_person_phonetic02 for > db:midoco.crm_person_phonetic. > > I assumed a charsset problem and made sure the DB_LOCALE was correct. > But data was not corrupt, and also oncheck -cD did not produce any > errors. > > After drop and re-creation of the index oncheck -cI gave the same output > ! > i_crm_person_phonetic02 is an index over the field forename, which > contains uppercase forenames. > > I also tested with 11.50UC8IE, same results. > > In my personal 11.50UC5DE (Developer Edition, also Ubuntu 32bit) the > problem does not occur. > Also in our production environment, the problem does not occur > (Workgroup Edition 64bit, 11.10FC3). > Thats where the data comes from, but should not make any difference, > since we only work on unload data. > > So I tried to find out if the error is occurring on that machine. But > the load with a copy of my Developer edition was OK. > I then tried to start up the dbspace from my developer edition with the > 11.50UC8IE (before, I did a oncheck -cI which was without errors). > After startup of the Innovator Edition (no data changes in between, but > internal conversion occurred), oncheck -cI for the same table > produced again the above results. > > Might be a bug in Innovator C ? > > Any hints ? > Marcus Haarmann > Geschäftsführer > > Midoco GmbH > Otto-Hahn-Str. 12 > 40721 Hilden > > Tel. +49 (2103) 28 74 0 > Fax. +49 (2103) 28 74 28 > www.midoco.de > > Member of Pisano Holding GmbH > Better Travel Technology > www.pisano-holding.com > > Amtsgericht Düsseldorf - HRB 51420 - USt.ID DE814319276 - > Geschäftsführer: Steffen Faradi, Marcus Haarmann, Jörg Hauschild - > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --bcaec501631feeb2aa049e9c9fd9

↪ replying to Art Kagel

Fernando Nunes — Wed, 16 Mar 2011 22:00:17 — source: IIUG Forums & Mailing Lists

Not enough details. It may be a bug in the engine. Include the table schema (including indexes) and your database locale in the PMR information. Regards. On Wed, Mar 16, 2011 at 5:10 PM, Art Kagel <art.kagel@gmail.com> wrote: > Sounds like a bug in oncheck. Call IBM and open a case. > > Art > > Art S. Kagel > Advanced DataTools (www.advancedatatools.com) > Blog: http://informix-myview.blogspot.com/ > > Disclaimer: Please keep in mind that my own opinions are my own opinions > and > do not reflect on my employer, Advanced DataTools, the IIUG, nor any other > organization with which I am associated either explicitly, implicitly, or > by > inference. Neither do those opinions reflect those of other individuals > affiliated with any entity with which I am affiliated nor those of the > entities themselves. > > On Wed, Mar 16, 2011 at 12:34 PM, Marcus Haarmann < > marcus.haarmann@midoco.de > > wrote: > > > Hi experts, > > > > we found an interesting problem while testing with some copied databases > > from out production system. > > Following situation: > > We used an instance of 11.70UC1IE on Linux 32 bit (Ubuntu 10.04) to load > > some data from production databases > > (using create table, load from, create index, trigger, foreign keys, > > done via dbaccess). > > > > We encountered the problem, that after the creation of the indexes, > > oncheck -cI had errors (freshly loaded, index created new). > > oncheck -cI derpart_eigen:crm_person_phonetic > > Validating indexes for db:midoco.crm_person_phonetic...> > Index 111_35 > > > > Index fragment partition rootdbs in DBspace rootdbs > > Index i_crm_person_phonetic01 > > > > Index fragment partition rootdbs in DBspace rootdbs > > Index i_crm_person_phonetic02 > > > > Index fragment partition rootdbs in DBspace rootdbs > > ERROR: Keys in node 0x2 are not in order. > > ERROR: Index i_crm_person_phonetic02 for db:midoco.crm_person_phonetic > > is bad. > > Index i_crm_person_phonetic03 > > Please Drop and ReCreate Index i_crm_person_phonetic02 for > > db:midoco.crm_person_phonetic. > > > > I assumed a charsset problem and made sure the DB_LOCALE was correct. > > But data was not corrupt, and also oncheck -cD did not produce any > > errors. > > > > After drop and re-creation of the index oncheck -cI gave the same output > > ! > > i_crm_person_phonetic02 is an index over the field forename, which > > contains uppercase forenames. > > > > I also tested with 11.50UC8IE, same results. > > > > In my personal 11.50UC5DE (Developer Edition, also Ubuntu 32bit) the > > problem does not occur. > > Also in our production environment, the problem does not occur > > (Workgroup Edition 64bit, 11.10FC3). > > Thats where the data comes from, but should not make any difference, > > since we only work on unload data. > > > > So I tried to find out if the error is occurring on that machine. But > > the load with a copy of my Developer edition was OK. > > I then tried to start up the dbspace from my developer edition with the > > 11.50UC8IE (before, I did a oncheck -cI which was without errors). > > After startup of the Innovator Edition (no data changes in between, but > > internal conversion occurred), oncheck -cI for the same table > > produced again the above results. > > > > Might be a bug in Innovator C ? > > > > Any hints ? > > Marcus Haarmann > > Geschäftsführer > > > > Midoco GmbH > > Otto-Hahn-Str. 12 > > 40721 Hilden > > > > Tel. +49 (2103) 28 74 0 > > Fax. +49 (2103) 28 74 28 > > www.midoco.de > > > > Member of Pisano Holding GmbH > > Better Travel Technology > > www.pisano-holding.com > > > > Amtsgericht Düsseldorf - HRB 51420 - USt.ID DE814319276 - > > Geschäftsführer: Steffen Faradi, Marcus Haarmann, Jörg Hauschild - > > > > > > > > > > ******************************************************************************* > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > --bcaec501631feeb2aa049e9c9fd9 > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently... --0015174c396cbd3bce049ea0aab2

↪ replying to Fernando Nunes

Marcus Haarmann — Thu, 17 Mar 2011 08:40:04 — source: IIUG Forums & Mailing Lists

Hi, I have tried to make a small test: drop table test; create table test ( forename nvarchar(10) ); load from forename.unl insert into test; create index i_test_forename on test (forename) using btree ; Locale is de_de.8859-15 I could reproduce this by loading 3000 rows in this table, containing german umlaut characters, empty values. dbaccess test test.sql Database selected. Table dropped. Table created. 3000 row(s) loaded. Index created. Database closed. root@janus:/srv# oncheck -cI test:test Validating indexes for test:root.test... Index i_test_forename Index fragment partition rootdbs in DBspace rootdbs ERROR: Keys in node 0x1 are not in order. ERROR: Index i_test_forename for test:root.test is bad. Please Drop and ReCreate Index i_test_forename for test:root.test. I could NOT reproduce this be loading less rows into the table, not depending if I throw out the first 1000 lines or the last 1000 lines. Strange. Marcus Haarmann Geschäftsführer Midoco GmbH Otto-Hahn-Str. 12 40721 Hilden Tel. +49 (2103) 28 74 0 Fax. +49 (2103) 28 74 28 www.midoco.de Member of Pisano Holding GmbH Better Travel Technology www.pisano-holding.com Amtsgericht Düsseldorf - HRB 51420 - USt.ID DE814319276 - Geschäftsführer: Steffen Faradi, Marcus Haarmann, Jörg Hauschild - -----Original Message----- From: Fernando Nunes [mailto:domusonline@gmail.com] Sent: Wednesday, March 16, 2011 11:00 PM To: ids@iiug.org Subject: Re: InnovatorC Bug ? [23124] Not enough details. It may be a bug in the engine. Include the table schema (including indexes) and your database locale in the PMR information. Regards. On Wed, Mar 16, 2011 at 5:10 PM, Art Kagel <art.kagel@gmail.com> wrote: > Sounds like a bug in oncheck. Call IBM and open a case. > > Art > > Art S. Kagel > Advanced DataTools (www.advancedatatools.com) > Blog: http://informix-myview.blogspot.com/ > > Disclaimer: Please keep in mind that my own opinions are my own > opinions and do not reflect on my employer, Advanced DataTools, the > IIUG, nor any other organization with which I am associated either > explicitly, implicitly, or by inference. Neither do those opinions > reflect those of other individuals affiliated with any entity with > which I am affiliated nor those of the entities themselves. > > On Wed, Mar 16, 2011 at 12:34 PM, Marcus Haarmann < > marcus.haarmann@midoco.de > > wrote: > > > Hi experts, > > > > we found an interesting problem while testing with some copied > > databases from out production system. > > Following situation: > > We used an instance of 11.70UC1IE on Linux 32 bit (Ubuntu 10.04) to > > load some data from production databases (using create table, load > > from, create index, trigger, foreign keys, done via dbaccess). > > > > We encountered the problem, that after the creation of the indexes, > > oncheck -cI had errors (freshly loaded, index created new). > > oncheck -cI derpart_eigen:crm_person_phonetic Validating indexes for > > db:midoco.crm_person_phonetic... > > Index 111_35 > > > > Index fragment partition rootdbs in DBspace rootdbs Index > > i_crm_person_phonetic01 > > > > Index fragment partition rootdbs in DBspace rootdbs Index > > i_crm_person_phonetic02 > > > > Index fragment partition rootdbs in DBspace rootdbs > > ERROR: Keys in node 0x2 are not in order. > > ERROR: Index i_crm_person_phonetic02 for > > db:midoco.crm_person_phonetic is bad. > > Index i_crm_person_phonetic03 > > Please Drop and ReCreate Index i_crm_person_phonetic02 for > > db:midoco.crm_person_phonetic. > > > > I assumed a charsset problem and made sure the DB_LOCALE was correct. > > But data was not corrupt, and also oncheck -cD did not produce any > > errors. > > > > After drop and re-creation of the index oncheck -cI gave the same > > output ! > > i_crm_person_phonetic02 is an index over the field forename, which > > contains uppercase forenames. > > > > I also tested with 11.50UC8IE, same results. > > > > In my personal 11.50UC5DE (Developer Edition, also Ubuntu 32bit) the > > problem does not occur. > > Also in our production environment, the problem does not occur > > (Workgroup Edition 64bit, 11.10FC3). > > Thats where the data comes from, but should not make any difference, > > since we only work on unload data. > > > > So I tried to find out if the error is occurring on that machine. > > But the load with a copy of my Developer edition was OK. > > I then tried to start up the dbspace from my developer edition with > > the 11.50UC8IE (before, I did a oncheck -cI which was without errors). > > After startup of the Innovator Edition (no data changes in between, > > but internal conversion occurred), oncheck -cI for the same table > > produced again the above results. > > > > Might be a bug in Innovator C ? > > > > Any hints ? > > Marcus Haarmann > > Geschftsfhrer > > > > Midoco GmbH > > Otto-Hahn-Str. 12 > > 40721 Hilden > > > > Tel. +49 (2103) 28 74 0 > > Fax. +49 (2103) 28 74 28 > > www.midoco.de > > > > Member of Pisano Holding GmbH > > Better Travel Technology > > www.pisano-holding.com > > > > Amtsgericht Dsseldorf - HRB 51420 - USt.ID DE814319276 - > > Geschftsfhrer: Steffen Faradi, Marcus Haarmann, Jrg Hauschild - > > > > > > > > > > ************************************************************************ ******* > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > --bcaec501631feeb2aa049e9c9fd9 > > > > ************************************************************************ ******* > Forum Note: Use "Reply" to post a response in the discussion forum. > > -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently... --0015174c396cbd3bce049ea0aab2 ************************************************************************ ******* Forum Note: Use "Reply" to post a response in the discussion forum.

InnovatorC Bug ?

Glossary

Your privacy choices