Spurious data in 'uninteresting' lines can get accidentally parsed by the getsummary script if they happen to match one of the patterns.
In my instance, the following line was in our results file:
[mount: none on /dev type tmpfs (rw,size=64k,mode=755)]
which I'm guessing matches the rule
if (/size=64/) {
in the getsummary script. This causes a whole slew of data to get ignored for that results file, ending up with only sparse reporting for that run.
As a quick fix I added the following rule near the top of the getsummary parser loop.
....
if (/^\[mount:/) {
next
}
if (/^\[lmbench/) {
....
This fixes my particular case, but is pretty horrible and I'm no perl expert, so I'm hoping somebody can come up with a much better perlesque fix.
Logged In: NO
Oops, needed to add email - you can decipher it...
graham tod whaley ta imgtec tod com
Graham