DB2 and SRDF

redwolf · #1 (**permalink**) 05-15-09, 10:40

We had an incident recently where data became corrupt at our R1 site due to a bad HBA card. Our UNIX team insisted that R2 would be corrupt also. We never checked our R2 data, but another team discovered their R2 data was not corrupted. Can someone explain how this worked that R1 got corrupted but R2 did not? I expected only data that was written to R1 would be captured and transmitted to R2. Would the data be transmitted to R2 before the R1 HBA corrupted it?

db2dummy1 · #2 (**permalink**) 05-23-09, 23:50

The only reason it appeared that the data is corrupt at R1 is because of the bad HBA. This was not an issue at R2 because all disks were accessible. DB2 went through crash recovery using logs, which were also moved by SRDF, and went on its way without incident.

Marcus_A · #3 (**permalink**) 05-24-09, 02:18

Quote:

Originally Posted by redwolf

We had an incident recently where data became corrupt at our R1 site due to a bad HBA card. Our UNIX team insisted that R2 would be corrupt also. We never checked our R2 data, but another team discovered their R2 data was not corrupted. Can someone explain how this worked that R1 got corrupted but R2 did not? I expected only data that was written to R1 would be captured and transmitted to R2. Would the data be transmitted to R2 before the R1 HBA corrupted it?

Can you explain what you mean by R1 and R2? I don't think these are DB2 terms.