WEBVTT

00:00:00.240 --> 00:00:00.760
[Autogenerated] requirement.

00:00:00.760 --> 00:00:03.450
10 men all about logging on monitoring.

00:00:03.450 --> 00:00:08.270
So we've we've really moved away now from preventative or protective controls

00:00:08.280 --> 00:00:10.970
and moving into starting to having some detective controls.

00:00:10.980 --> 00:00:15.520
Only because the idea of requirement 10 is that you

00:00:15.520 --> 00:00:17.730
confined bad things that are happening.

00:00:17.740 --> 00:00:20.650
Yes, in real time or very nearly.

00:00:20.660 --> 00:00:22.520
Yeah, it's no on.

00:00:22.520 --> 00:00:23.860
This is what people keep.

00:00:23.870 --> 00:00:25.530
A lot of people say to me,

00:00:25.540 --> 00:00:28.040
you but requirement tenets only really useful if we get breached,

00:00:28.040 --> 00:00:28.300
isn't it?

00:00:28.300 --> 00:00:29.750
That's the only time we need our logs.

00:00:30.140 --> 00:00:35.850
Maybe you can make a breached less severe right by catching it sooner.

00:00:35.860 --> 00:00:36.280
Right?

00:00:36.290 --> 00:00:40.910
Or or maybe you can sort of catch it in progress and

00:00:40.920 --> 00:00:43.110
prevent something bad from happening.

00:00:43.120 --> 00:00:43.940
Uh, you know,

00:00:43.940 --> 00:00:46.970
if you go look at any of the statistics about actual security

00:00:46.970 --> 00:00:48.900
incidents like from the Verizon data breaches,

00:00:48.910 --> 00:00:53.080
uh, investigation report right when they when they look at the time scale,

00:00:53.080 --> 00:00:53.290
right.

00:00:53.290 --> 00:00:57.270
The order of magnitude of time for the attacker to make first

00:00:57.270 --> 00:01:00.590
contact for the attacker to start XSLT rating cardholder data for

00:01:00.600 --> 00:01:03.340
the organization to detect what has happened,

00:01:03.340 --> 00:01:03.910
right, That's that's,

00:01:03.910 --> 00:01:08.550
like seconds or minutes and then minutes or hours and then weeks or months.

00:01:08.560 --> 00:01:11.890
Uh, because they're not really looking at their logs.

00:01:11.900 --> 00:01:12.270
Right?

00:01:12.270 --> 00:01:15.860
Or they're they're doing something very perfunctory and very sort of check box.

00:01:15.840 --> 00:01:18.310
Uh, that doesn't That doesn't tell them anything.

00:01:18.320 --> 00:01:23.640
Yeah, and this is one thing that I struggle to describe to lots of people.

00:01:23.640 --> 00:01:24.120
Really.

00:01:24.130 --> 00:01:26.550
DSS is also about detection.

00:01:26.550 --> 00:01:27.990
It's about alright.

00:01:27.990 --> 00:01:31.420
It's only just one is one requirement out of 12, right?

00:01:31.430 --> 00:01:32.600
But it's about you.

00:01:32.610 --> 00:01:32.910
You.

00:01:32.920 --> 00:01:35.200
There's something not right in your network or something,

00:01:35.200 --> 00:01:36.230
not right in your systems.

00:01:36.230 --> 00:01:37.850
You've got the opportunity to spot it.

00:01:37.860 --> 00:01:40.040
E from Verizon.

00:01:40.040 --> 00:01:42.170
You know, it is weeks or months in that database report.

00:01:42.170 --> 00:01:44.990
It's generally averages in, like, 100 days or something like that.

00:01:45.000 --> 00:01:45.860
A lot.

00:01:45.840 --> 00:01:46.900
I don't know if you noticed.

00:01:46.900 --> 00:01:50.020
Recently there was a large compromise of the European airline,

00:01:50.030 --> 00:01:52.730
Yes, and that the detection papers 15 days,

00:01:52.730 --> 00:01:54.230
and I was thinking that was pretty good.

00:01:54.230 --> 00:01:56.690
It's pretty good on all of the Internet went into meltdown,

00:01:56.690 --> 00:01:58.920
saying, Well, how come it took 15 days and I'm thinking,

00:01:58.920 --> 00:02:00.540
Wow, I thought 15 days was good,

00:02:00.550 --> 00:02:04.510
but only thinks the internet minutes to forget about what happened before.

00:02:04.520 --> 00:02:09.130
So logging and monitoring We'll talk about that first,

00:02:09.130 --> 00:02:12.400
and then we'll talk about time because time waits for no one.

00:02:12.410 --> 00:02:12.860
Right.

00:02:13.540 --> 00:02:17.810
So you have to buy something to do this properly now, don't you?

00:02:17.820 --> 00:02:23.450
E I don't think you could do it without without at least some some tools, right?

00:02:23.450 --> 00:02:26.690
Some of them are open source, but they take a little bit more sort of effort.

00:02:26.700 --> 00:02:28.830
But yeah, the idea that you could just, like,

00:02:28.840 --> 00:02:33.100
task an intern to, like, watch the logline scroll by is ridiculous,

00:02:33.110 --> 00:02:33.700
right?

00:02:33.710 --> 00:02:34.180
S.

00:02:34.180 --> 00:02:34.370
Oh, yeah.

00:02:34.370 --> 00:02:38.880
You need something to collect all that log data to do some analytics, right?

00:02:38.880 --> 00:02:43.570
To filter out all of the sort of boring stuff and bring

00:02:43.570 --> 00:02:45.820
to the surface the interesting things, right?

00:02:45.830 --> 00:02:48.660
So it's a que se you walk on site,

00:02:48.670 --> 00:02:51.580
you You're going to do an assessment of requirement

00:02:51.590 --> 00:02:53.170
that What are you looking for?

00:02:53.180 --> 00:02:58.660
Eso The two pieces are Show me what data you are collecting in the first place,

00:02:58.670 --> 00:02:58.970
right?

00:02:58.980 --> 00:03:01.070
All the stuff in 10.2 in 10.3.

00:03:01.040 --> 00:03:02.700
Show me how you're doing this.

00:03:02.710 --> 00:03:03.170
Okay?

00:03:03.140 --> 00:03:03.640
How are you?

00:03:03.640 --> 00:03:06.490
How are you logging this at the appropriate layer?

00:03:06.500 --> 00:03:06.850
Right?

00:03:06.860 --> 00:03:07.480
Right on.

00:03:07.480 --> 00:03:09.820
In some cases, that's gonna be the operating system logs.

00:03:09.830 --> 00:03:10.430
Right?

00:03:10.440 --> 00:03:13.440
Uh, tell me about logging failures at the operating system.

00:03:13.440 --> 00:03:13.610
Right.

00:03:13.610 --> 00:03:17.020
If if an administrator user account fails,

00:03:17.030 --> 00:03:19.520
you know Loggins 10 times in a row at three.

00:03:19.520 --> 00:03:20.280
In the morning.

00:03:20.290 --> 00:03:22.670
You want to know that, right?

00:03:22.640 --> 00:03:25.400
And sometimes is that at an application layer, right?

00:03:25.400 --> 00:03:27.210
And if they're developing their own application,

00:03:27.210 --> 00:03:30.680
they have to develop its capability to log these events.

00:03:30.690 --> 00:03:31.960
So, what are you looking at?

00:03:31.960 --> 00:03:32.390
A necessity.

00:03:32.390 --> 00:03:33.900
You're looking at the logs you actually trying to

00:03:33.900 --> 00:03:36.850
create you trying to You're saying, looking in the wrong password.

00:03:36.860 --> 00:03:37.920
You know,

00:03:37.920 --> 00:03:41.760
I generally don't force a client to create a

00:03:41.760 --> 00:03:44.910
situation just to produce a log event.

00:03:44.920 --> 00:03:48.720
I mean, maybe something relatively low key, like flub the password.

00:03:48.720 --> 00:03:52.410
And then let's see the log, you know, is not is not a big ask,

00:03:52.420 --> 00:03:55.750
but, you know, go delete a system level object.

00:03:55.740 --> 00:03:56.170
URL IT.

00:03:56.170 --> 00:03:58.930
Go drop a table in your database, just like yeah,

00:03:58.940 --> 00:04:00.570
that's that's maybe a bridge too far.

00:04:01.240 --> 00:04:07.050
But I do want to see yes, some log events that show these these criteria right?

00:04:07.050 --> 00:04:09.000
Described in 10.2 in 10.3.

00:04:09.010 --> 00:04:10.160
Um, So, yeah,

00:04:10.170 --> 00:04:14.550
go find those for us or go give us some sample of data and then

00:04:14.550 --> 00:04:16.329
we'll look for them in that in that spread.
