Sorry for my long silence too (also buzy...). We have done a few tests, everything with FB 2.1.4.
The problem was reproducible: each restore (from the one and the same backup) produced the same problem. (BTW, we are regularly updating this DB/table weekly. After each backup/restore we are getting the same error.) Then I did the following: On machine A... - dropped all other tables - dropped the primary index - backup On machine B... - restore - validate => OK - created 6 indices, each based on a single field from the previous primary index * validate => all indices with en error - at this point I thought: "If I were Vlad, I would never believe this story", so I was not in a mood to write anything Then we thought: OK, let us see, what can we do with a DB, in order to give it to Vlad :) So, - drop all indices - drop all fields apart form those 6 - drop even 3 of these 6 fields (so that we were left with a table with only 6 fields, hoping very much not to loose the error - the reason was, that we would have no problem with giving their content away) - create index (am not sure if that was one on all 3 fields, or 3 separate indices) - validate => no error (damn it!) - at this point I thought even more: "If I were Vlad, I would never believe this story!", so I was again not in a mood to write anything. Now to your questions: > So, all key fileds is VARCHAR with charset NONE, correct ? Yes. > Is it possible to have zero bytes in data values of any of the key fields ? No, but no systematic test was made. However, I do think we have a routine that eliminates zero bytes from the text material before reading it in. So, I would say again: no, no zero bytes. > What about leading or trailing zero's ? Leading zeros (character 0, not zero byte) are extremely likely in some of the fields (exactly those we deleted from the final tests). >> I know that this is not recommended. I was already >> thinking of defining it to be UTF-8 and than making backup and restore. We have not done these tests yet, sorry. >> Also, I was today thinking of defining 6 indices, each based on only one >> field of the current primary key and see what happens. >> >> Are these tests worth trying? > > Sure. Have described these results above. >>> Is it reproducible ? I.e. if you restore same backup few times it >>> have missed entries every time ? Yes, that was the case, as described above. So, this is the current state of affairs. I shall make tests with other key fields (those that, when dropped, made the error wanish). However, it will last a while, so please have patience. Thank you. Regards, Borut ------------------------------------------------------------------------------ WhatsUp Gold - Download Free Network Management Software The most intuitive, comprehensive, and cost-effective network management toolset available today. Delivers lowest initial acquisition cost and overall TCO of any competing solution. http://p.sf.net/sfu/whatsupgold-sd Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel