J
Joe Faulhaber[MSFT]
Some answers for you, Alan:
Those zillions of Defender checkpoints still aren't completely understood by
us, but we've made a bunch of changes in house that should fix them for our
next set of bits...we're going to only do one checkpoint a day (the released
bits max is one/hour), and they should ONLY be snapped before cleaning
threats, right now they're snapped when taking any action on unknowns or
threats, and we're still not sure how scanning is creating checkpoints. So
those are lame, and we're going to change them. We have a KB on how to turn
the darn things off from WD, I can't find it right now, I'll look it up.
The bad guys have written a bunch of tools so when patch Tuesday rolls
around, they can quickly write exploits for whatever just got fixed. So
we're tending to see exploits come out for many vulnerabilities only _after_
the patch is issued. And then we're in a race to update everybody before
the bad guys get to unpatched machines. It's a darn tough problem to solve,
too - of course not having the vulnerability in the first place is the best
fix.
Regards,
Joe
Those zillions of Defender checkpoints still aren't completely understood by
us, but we've made a bunch of changes in house that should fix them for our
next set of bits...we're going to only do one checkpoint a day (the released
bits max is one/hour), and they should ONLY be snapped before cleaning
threats, right now they're snapped when taking any action on unknowns or
threats, and we're still not sure how scanning is creating checkpoints. So
those are lame, and we're going to change them. We have a KB on how to turn
the darn things off from WD, I can't find it right now, I'll look it up.
The bad guys have written a bunch of tools so when patch Tuesday rolls
around, they can quickly write exploits for whatever just got fixed. So
we're tending to see exploits come out for many vulnerabilities only _after_
the patch is issued. And then we're in a race to update everybody before
the bad guys get to unpatched machines. It's a darn tough problem to solve,
too - of course not having the vulnerability in the first place is the best
fix.
Regards,
Joe