Steve Brett > hello all, > > I've created a database with 50 million random people with > random addresses > so we can test sapdb and use it in future projects. > > the data was uploaded using the replication manager and i > started on friday > and came back today (long weekend in uk for bank holiday) to > find that one > of the link tables had failed. this was down to my own error > (doh!) but left > me in the position of having the recreate the table. > > i couldn't re-run the program i worte to generate the data as > it was all > randomly created and the link table needs to describe a > relationship between > two tables so i did the following: > > sql_execute create table patient_cnames(patient_id,cname_id,primary > key(patient_id,cname_id)) as select p.patient_id,fn.fname_id > from patient > p,fnames fn, patient_fnames pfn where p.patient_id=pfn.fnames_id and > pfn.fname_id=fn.fname_id and pfn.position=1 > > this ran for about 2 hours and then the following was output: > > error: > > ERR > -24988,ERR_SQL: sql error > 500,Lock request timeout > > > does anyone know whay this might have happenned ? i've > searched the archives > but not found anything similar .... > > thanks,
Usually error 500 will be returned if there are at least two tasks/sessions working on the same table(s). In your case this may happen if some else does an insert/update/delete on table patient, fnames or patient_fnames and holds the corresponding exclusive lock longer than your installation parameter REQUEST_TIMEOUT (given in seconds) allows YOU to wait. To avoid this, isolation level 0 (dirty read for your select) or isolation level 2 or 3 (table lock, no parallel DML possible, parallel select possible) could be used. Elke SAP Labs Berlin _______________________________________________ sapdb.general mailing list [EMAIL PROTECTED] http://listserv.sap.com/mailman/listinfo/sapdb.general
