- Article Type: General
- Product: Aleph
- Product Version: 20, 21, 22, 23
ue_01_word_parallel stops with the following error (in the $data_scratch/run_e_01_word.nnnn log):
Execution error : file '/exlibris/aleph/a22_1/aleph/exe/io_word3_a.gnt'
error code: 115, pc=0, call=1, seg=0
115 Unexpected signal (Signal 6)
(Note the "115".)
In another case -- the $data_scratch/run_e_01_z0102 log for ue_01_z0102 -- the following was seen:
Execution error : file '/exlibris/aleph/a17_1/aleph/exe/io_word3_a.gnt'
error code: 114, pc=0, call=57, seg=0
114 Attempt to access item beyond bounds of memory (Signal 11)
(Note the "114".)
To diagnose such a problem, you need to record the bib number (shown in the "HANDLING DOC NO" line immediately preceding the error) and re-send the record to the server to see if it results in the error again.
If it does, please consider what might be unique about the record. If there's something different about the record, then find other records with these same characteristics and try sending them to the server to see if *they* get the error.
If the problem happens with multiple records, consider whether the Execution error is preceded by a line like this:
Update z98 from z980 : 019000760295
If the 12-digit number (the z980 number) is the same in each case, it may be that there's something wrong with this z980. In this case, there had been a problem with bib records being incorrectly loaded with doc numbers higher than the util g/2 "LAST-DOC-NUMBER". And the following SQL showed a z980 with a z980_no_entries value of 92:
SQL> select z980_rec_key, z980_no_entries from z980 where z980_no_entries > '90';
(The maximum is 90; anything above is an error.) The following SQL was used to delete this z980 record:
SQL> delete from from z980 where z980_rec_key = '019000760295';
In the above case, the word for this z980 word number was the $$h time in the CAT field, which would hardly ever be searched, so deleting this z980 was not a problem. The article " Finding z97 word from word number " describes how to see what the word is.
If the word is a common one, then the solution would be to run manage-01 to recreate the Word indexes. See the article Aleph Indexing Methods for the best way to do this.
In the other case (the run_e_01_z0102), the problem did *not* occur when the bib record was resent to the server.
- Article last edited: 12-Jan-2017