G
gus bjorklund
Guest
hard to know for sure. most likely not. but: when the error 5028 and 2522 were detected, the detector (WDOG in this case) would have turned on the abnormal shutdown flag. this flag is checked in the relevant write functions before all data extent, bi, and ai writes and if enabled then the write is skipped. this applies to all database connections, self-serving, served, and utility. apw, biw and aiw too. when a process discovers that the abnormal shutdown flag is enabled, it is supposed to go away as soon as possible, without doing anything more. but it may be connected to other databases from which it has to back out transactions. often, when this sitation arises, a variety of side-effects occur with error messages. like the error 1007 from the WDOG. it is possible that the begin transaction backout messages were written but the transactions were not actually backed out in the database in question because all database i/o was disabled. the code that writes the messages does not discover that until after the messages are written. to answer the question: it should not have been possible to restart the database while there were old processes still connected. the startup code has checks to see if shared memory exists and if there are processes with old extant shared memory attachments before starting anything. if so, that code is supposed to prevent startup. without examining the bi file it is impossible to know if something got written by an old process from the previous session. an ai scan might reveal something. -gus
Continue reading...
Continue reading...