Re: Informix disk mirroring -- what good is it?

Interesting... Page cleaners also die when a controller is swamped by device requests. Are you sure the disk actually went bad? Informix recommends limiting the number of page cleaners based on the number of disk controllers. What release of 5.0 were you running (tbstat -V)? 5.04.UC1? How many buffers, LRU Qs (tbstat -R)? How many page cleaners (tbstat -F)? How many disk controllers? How frequent were your checkpoints? Were you using buffered logging? Thanks. In article <3ueg3n$5qg@host-04.colby.edu>, dwcooley@colby.edu (David W. Cooley) wrote: > We are running INFORMIX/Online 5.0 with INFORMIX based mirroring on an HP 9000 > G40 running HP-UX 9.04. > > When we initially installed the software we discussed the tradeoffs of HP-UX LVM based mirroring and the INFORMIX database engine based mirroring. We explored this issue with our application vendor and with INFORMIX. Both stated unequivically that we should use the INFORMIX based mirroring because the database engine has a much better understanding of the data than the operating system does and thus would optimize i/o and would processes transactions correctly (there was a strong implication that o/s based mirroring would lead to > data corruption). So we went with the advice but skeptically. > > Further conversations with a peer institution revealed that INFORMIX based mirroring did not always handle disk problems correctly. The contact reported that they had one experience where a disk died and the database engine was unable to continue operations. Followup by the institution with both their application vendor and INFORMIX was unable to indentify why the database engine was unable to apropriately handle the situation. > > Last week we had a similar experience. The disk on which a mirror chunck resides died. The database engine was unable to detect that the disk was unavailable except that the page cleaners all timed out. The database engine locked up and would not respond to any commands including shutdown. The only > recourse was to reboot the system. The system marked the mirror down and preceded to restore consistancy. Unfortunately, its idea of consistancy was not > consistant enough for normal operations and the database engine complained about missing rowid's. > > Our main objective in running mirroring is to ensure that the database environment will survive a disk failure and that operations will continue uninterrupted. Obviously this objective is not being met. > > The next step for us is to question the usefullness of INFORMIX based mirroring and seek alternative ways of accomplishing the same objective. > > Do others have any experiences or information on this issue which might shed some light on the matter? > > -Dave > > > -- > David W. Cooley e-mail: dwcooley@colby.edu > Associate Director for Technical Services phone: (207)872-3581 > Computer Services fax: (207)872-3555 > Colby College > Waterville, Maine 04901 -- charlie_sadler@mathworks.com