WEBVTT

00:00:01.340 --> 00:00:04.700
[Autogenerated] requirement 10.6 now wants you to do something with

00:00:04.700 --> 00:00:06.350
all the data that's being written to the logs.

00:00:06.740 --> 00:00:11.020
There's something it wants you to do is obviously to review them on again.

00:00:11.110 --> 00:00:14.030
DSS is showing its age here because the note to the

00:00:14.030 --> 00:00:17.180
requirements suggests that log harvesting passing on alerting

00:00:17.180 --> 00:00:19.780
tools may be used to meet this requirement.

00:00:19.960 --> 00:00:21.570
You know, in the current decade,

00:00:21.570 --> 00:00:24.460
where log files could be gigabytes sized very quickly.

00:00:24.790 --> 00:00:28.550
But the concept of not using automated TLS is pretty inconceivable.

00:00:29.440 --> 00:00:32.870
There isn't an explicit testing procedure for 10.6 the tests error.

00:00:32.870 --> 00:00:35.930
Giving for each of the sub requirements on the first of

00:00:35.930 --> 00:00:38.890
requirement is that the important logs and events should be

00:00:38.890 --> 00:00:40.450
reviewed at least once a day.

00:00:40.940 --> 00:00:43.370
The next requirement going to look at is that less important.

00:00:43.370 --> 00:00:46.250
Logs and events review less frequently than one us today.

00:00:47.140 --> 00:00:50.750
Now splitting log reviews in to what's important and what's less

00:00:50.750 --> 00:00:53.650
important was quite a recent change in the standard.

00:00:53.900 --> 00:00:56.790
It came in response to how people typically round there

00:00:56.790 --> 00:00:59.350
seems on their security operations teams.

00:01:00.040 --> 00:01:03.730
It's important to understand the intent of this requirement here,

00:01:03.870 --> 00:01:07.890
which is the operations team should be able to prioritize important events

00:01:08.100 --> 00:01:11.050
that may indicate an attack against card all the data,

00:01:11.540 --> 00:01:16.070
so a minimum on a daily basis, these four things need to be reviewed.

00:01:16.440 --> 00:01:18.170
Firstly, any security events.

00:01:18.450 --> 00:01:23.160
Secondly, logs from systems that store process or transmit cardholder data.

00:01:23.740 --> 00:01:29.070
Thirdly, blog from critical systems such as directory servers and finally,

00:01:29.200 --> 00:01:33.200
logs from any component that exist to provide a security function like a

00:01:33.210 --> 00:01:37.800
firewall on to test this requirement is being fulfilled.

00:01:37.970 --> 00:01:41.750
The QoS A will check that the log reviews are required by a policy

00:01:42.540 --> 00:01:45.610
and will interview people responsible for log reviews and event

00:01:45.610 --> 00:01:48.060
response and also watch them working.

00:01:49.240 --> 00:01:49.530
So,

00:01:49.540 --> 00:01:52.020
having split the requirement to categorize logs and

00:01:52.020 --> 00:01:55.340
events as important and less important, of course,

00:01:55.340 --> 00:01:57.850
those less important things also need to be reviewed,

00:01:59.540 --> 00:02:02.980
and how frequently need to be reviewed is up to the organization.

00:02:03.150 --> 00:02:06.010
The standard uses the term periodically,

00:02:06.150 --> 00:02:09.169
which I know really upset some people because they want the

00:02:09.169 --> 00:02:12.320
standard to tell them exactly how frequently these less

00:02:12.320 --> 00:02:13.850
important log should be reviewed.

00:02:14.440 --> 00:02:16.740
Although I find that these error also the people who say

00:02:16.740 --> 00:02:18.450
they want the standard be more is spaced,

00:02:18.700 --> 00:02:23.290
so look periodically equates to risk based how frequently logs

00:02:23.290 --> 00:02:26.290
for less important systems need to be viewed is based on the

00:02:26.290 --> 00:02:28.150
organization's risk assessment.

00:02:28.940 --> 00:02:33.530
And so the first thing that the US will check is that risk assessment to

00:02:33.530 --> 00:02:37.750
understand how frequently organization has decided periodic is.

00:02:38.340 --> 00:02:41.040
They will then want to look at the policies and procedures to

00:02:41.040 --> 00:02:43.980
see that those policies and procedures support the log reviews

00:02:44.090 --> 00:02:45.780
based on the risk assessment on.

00:02:45.780 --> 00:02:50.500
Finally, they'll talk with people who actually do the reviews unsurprising.

00:02:50.500 --> 00:02:52.520
The last sub, Rick want to of 10.6,

00:02:52.620 --> 00:02:55.670
is to actually follow up on things that the log reviews suggest.

00:02:55.670 --> 00:02:57.170
A not normal or expected.

00:02:57.640 --> 00:02:58.150
Okay,

00:02:58.220 --> 00:03:03.200
so the exact was the standard uses are follow upon anomalies and exceptions.

00:03:03.340 --> 00:03:06.550
But I guess most of us now think of a set of correlated log

00:03:06.550 --> 00:03:09.230
entries that generate an event for investigations,

00:03:09.230 --> 00:03:11.740
the sort of thing that we should look at in terms

00:03:11.740 --> 00:03:13.270
of how a modern soccer operate.

00:03:13.410 --> 00:03:15.050
This requirement is very basic,

00:03:15.540 --> 00:03:19.450
but the core principle that standard is looking for is that if someone or

00:03:19.450 --> 00:03:22.250
some system detects science of suspicious behavior,

00:03:22.350 --> 00:03:24.660
that this suspicious behaviors investigated,

00:03:25.340 --> 00:03:27.750
the testing procedures also pretty basic.

00:03:27.890 --> 00:03:30.410
It's not looking for evidence of event to reality.

00:03:30.410 --> 00:03:31.520
The log of actions taken,

00:03:31.520 --> 00:03:34.370
which you might expect normally in a modern day security

00:03:34.370 --> 00:03:37.530
operation center the accessible walk to first e.

00:03:37.530 --> 00:03:39.920
Check that there's a policy and procedures to follow up on

00:03:39.930 --> 00:03:42.070
anomalies and exceptions found in the logs.

00:03:42.440 --> 00:03:47.150
Then they'll interview people on observe them to see how events are followed up.
