Every time mongod or mongos crashes, we’ll see some hex numbers and the backtraces. The backtraces looks messy and ugly and needs to be parsed into human readable things.
I want to know in mongodb team, how do you make best use of the hex number list, and the backtrace which may contains hex or symbols to find the potential cause of the crashes ? In extreme situation, do we have to use some kind of disassembly techniques to explain those hex numbers ?
As far as I know, there’re two tools , c++filt, and addr2line. Any other methods ?