WEBVTT

0:00:01.820000 --> 0:00:07.300000
 In this video, we're going to take a
 look at configuring diagnostics on

0:00:07.300000 --> 0:00:14.080000
 resources. The video is going to be
 pretty much just a demonstration.

0:00:14.080000 --> 0:00:18.520000
 We're going to take a look at configuring
 virtual machine diagnostics,

0:00:18.520000 --> 0:00:22.560000
 and then configuring web app diagnostics,
 and then we're going to take

0:00:22.560000 --> 0:00:24.080000
 a look at other resources.

0:00:24.080000 --> 0:00:28.420000
 In this case, I'm going to demonstrate
 an Azure SQL Database and Azure

0:00:28.420000 --> 0:00:32.920000
 SQL Server. The point of this demonstration
 is to show you a few things.

0:00:32.920000 --> 0:00:38.960000
 First, that you can configure diagnostics
 and metric collection for different

0:00:38.960000 --> 0:00:45.220000
 resources. But more importantly, each
 resource has its own quirks when

0:00:45.220000 --> 0:00:50.280000
 it comes to configuring it
 for diagnostic purposes.

0:00:50.280000 --> 0:00:55.140000
 With that said, let's go ahead and take
 a look at the process of configuring

0:00:55.140000 --> 0:01:00.300000
 diagnostics. This is not a terribly
 lengthy demonstration.

0:01:00.300000 --> 0:01:05.040000
 Let's go ahead and
 pop that over here.

0:01:05.040000 --> 0:01:11.200000
 I have a resource group, and in the
 resource group, I've got all of the

0:01:11.200000 --> 0:01:14.340000
 resources that I want to
 set up diagnostics for.

0:01:14.340000 --> 0:01:20.760000
 I've got a virtual machine, I have a
 web app, and I have a SQL Database

0:01:20.760000 --> 0:01:22.900000
 and an Azure SQL Server.

0:01:22.900000 --> 0:01:29.600000
 I'm going to start out by going into
 my virtual machine configuration,

0:01:29.600000 --> 0:01:35.180000
 and I'm going to come down
 to Diagnostics Settings.

0:01:35.180000 --> 0:01:39.860000
 There's a few things
 that I want to do.

0:01:39.860000 --> 0:01:44.600000
 First of all, I want to go ahead and
 enable guest level diagnostics.

0:01:44.600000 --> 0:01:50.960000
 Now, with a virtual machine when you enable
 guest level diagnostics, what's

0:01:50.960000 --> 0:01:55.600000
 going to happen is that it's actually
 going to install an extension into

0:01:55.600000 --> 0:02:01.280000
 the virtual machine to collect
 extended diagnostic data.

0:02:01.280000 --> 0:02:08.200000
 Now, in addition to this, you will often
 want to set up the boot diagnostics.

0:02:08.200000 --> 0:02:12.140000
 I'm not sure if it's going to
 let me change it is indeed.

0:02:12.140000 --> 0:02:17.460000
 It's telling me that boot diagnostics
 is not configured for this virtual

0:02:17.460000 --> 0:02:22.640000
 machine. I'm going to go ahead and this
 is a very complex process, turn

0:02:22.640000 --> 0:02:24.520000
 on boot diagnostics.

0:02:24.520000 --> 0:02:28.040000
 Now, when you turn on boot diagnostics,
 those diagnostics need to go into

0:02:28.040000 --> 0:02:35.560000
 a storage account, and I'm going to go
 ahead and save my boot diagnostics.

0:02:35.560000 --> 0:02:41.240000
 Now, the boot diagnostics and the extended
 diagnostics both go into storage

0:02:41.240000 --> 0:02:45.280000
 accounts, and typically, you would
 most likely put them into the same

0:02:45.280000 --> 0:02:49.480000
 storage account, but I
 was doing this quickly.

0:02:49.480000 --> 0:02:54.120000
 I believe it's not going to, yeah,
 I'm going to have to wait here.

0:02:54.120000 --> 0:03:00.040000
 So, I'm going to wait, and for you, the
 video should be sped up to finish

0:03:00.040000 --> 0:03:11.200000
 this out. All right, we have now successfully
 set up diagnostics for our

0:03:11.200000 --> 0:03:15.240000
 virtual machine.

0:03:15.240000 --> 0:03:24.140000
 And here's my log of the last diagnostic,
 which it hadn't actually been

0:03:24.140000 --> 0:03:29.120000
 able to set up yet because it was rebooting
 while it was still being created.

0:03:29.120000 --> 0:03:36.780000
 And if I go to my diagnostic settings,
 now, in addition to seeing all

0:03:36.780000 --> 0:03:40.120000
 the information, because I turned that
 on, I actually have very detailed

0:03:40.120000 --> 0:03:44.080000
 capabilities in terms
 of what I can record.

0:03:44.080000 --> 0:03:47.020000
 So, I've got performance counters, and
 I've got basic level performance

0:03:47.020000 --> 0:03:52.600000
 counters, or if I go custom, I've got
 all of the performance counters

0:03:52.600000 --> 0:03:58.020000
 on the machine. There's pages of these,
 but we'll stick with basic logs.

0:03:58.020000 --> 0:04:02.080000
 Same idea. I've got various logs that
 I can collect, or I can customize

0:04:02.080000 --> 0:04:04.880000
 that and get even more detailed.

0:04:04.880000 --> 0:04:09.500000
 I have the choice of collecting
 crash dumps.

0:04:09.500000 --> 0:04:16.460000
 Now, with syncs, I can send this
 to application insights.

0:04:16.460000 --> 0:04:22.280000
 And then agent is the basic configuration
 for my diagnostics.

0:04:22.280000 --> 0:04:27.760000
 You can see how much disk space
 I'm willing to take up.

0:04:27.760000 --> 0:04:31.760000
 I can see the infrastructure, I want
 diagnostic infrastructure logs, my

0:04:31.760000 --> 0:04:33.560000
 log level, et cetera.

0:04:33.560000 --> 0:04:38.080000
 So, that is configuring logging
 for a virtual machine.

0:04:38.080000 --> 0:04:43.780000
 Next, let's take a look
 at setting up logging.

0:04:43.780000 --> 0:04:47.880000
 That's fine. I don't really want to
 make any changes there anyways.

0:04:47.880000 --> 0:04:53.240000
 Let's take a look at setting
 up logging on a web app.

0:04:53.240000 --> 0:04:56.020000
 There's my web app.

0:04:56.020000 --> 0:05:02.340000
 This is a little more straightforward,
 but it is a different setup for

0:05:02.340000 --> 0:05:09.140000
 logging. I'm going to come down to
 my diagnostic logs, and I can turn

0:05:09.140000 --> 0:05:14.140000
 them on. I have application logging
 to the file system and also to blob

0:05:14.140000 --> 0:05:18.140000
 storage. If I turn on application logging
 to the file system, that's going

0:05:18.140000 --> 0:05:23.520000
 to log to the running compute
 instances behind my web app.

0:05:23.520000 --> 0:05:28.620000
 And I can set the level that I want to
 log, maybe I'll log warnings there.

0:05:28.620000 --> 0:05:35.700000
 I can also separately log to blob storage,
 again, at the application level.

0:05:35.700000 --> 0:05:41.260000
 Maybe I just want to log errors, and
 I would have to set up my storage

0:05:41.260000 --> 0:05:46.680000
 settings. Here's a monitoring.

0:05:46.680000 --> 0:05:57.520000
 Let's just add web
 app as a container.

0:05:57.520000 --> 0:06:02.720000
 Select that container.

0:06:02.720000 --> 0:06:04.400000
 Specify retention period.

0:06:04.400000 --> 0:06:09.820000
 It's 365 days. Then I can
 also set up my web server.

0:06:09.820000 --> 0:06:12.280000
 This is, for example,
 at the IIS level.

0:06:12.280000 --> 0:06:15.640000
 It's a little different rather than
 being able to set the file system

0:06:15.640000 --> 0:06:19.440000
 and blob storage independently,
 I can choose one or the other.

0:06:19.440000 --> 0:06:23.240000
 Say I want to send my web
 server logging to storage.

0:06:23.240000 --> 0:06:26.480000
 Then I can set my storage account.


0:06:26.480000 --> 0:06:33.840000
 This one. Go ahead and send
 it to the same web app.

0:06:33.840000 --> 0:06:35.540000
 It's a retention day.

0:06:35.540000 --> 0:06:39.820000
 It's also 365. I also have the ability
 to have detailed error messages

0:06:39.820000 --> 0:06:42.260000
 and failed request tracing.

0:06:42.260000 --> 0:06:45.280000
 These are all the options
 for a web app.

0:06:45.280000 --> 0:06:51.340000
 Now, collecting data from the application
 separate from the IIS platform

0:06:51.340000 --> 0:06:55.700000
 or the web platform makes sense, right,
 because you have application errors.

0:06:55.700000 --> 0:07:01.140000
 Maybe somebody tries to upload data
 to a page that doesn't accept that

0:07:01.140000 --> 0:07:06.680000
 data. Whereas I also have
 my service level errors.

0:07:06.680000 --> 0:07:11.460000
 For example, maybe the service
 crashes for some reason.

0:07:11.460000 --> 0:07:16.180000
 Being able to track those separately
 does, in fact, make sense.

0:07:16.180000 --> 0:07:18.620000
 I can also download the logs.

0:07:18.620000 --> 0:07:23.000000
 I can see information on how to download
 my web application logs.

0:07:23.000000 --> 0:07:27.220000
 Then I can just save
 those settings.

0:07:27.220000 --> 0:07:32.880000
 Web applications can also be integrated
 with application insights.

0:07:32.880000 --> 0:07:34.380000
 I want to do that.

0:07:34.380000 --> 0:07:37.260000
 I'm not going to turn this on because
 this does take a little while.

0:07:37.260000 --> 0:07:41.960000
 I can turn on that site extension and
 further integrate a very rich set

0:07:41.960000 --> 0:07:45.640000
 of analytics for the
 application itself.

0:07:45.640000 --> 0:07:50.140000
 The next thing that I want to do, which
 pop over to databases here, I

0:07:50.140000 --> 0:07:58.120000
 have a SQL database and I'm going to
 set up my logging for this, which

0:07:58.120000 --> 0:08:02.260000
 is really more auditing
 than logging.

0:08:02.260000 --> 0:08:07.820000
 I have both. I'm going to go to the
 diagnostics of my SQL database.

0:08:07.820000 --> 0:08:12.140000
 I can turn on diagnostics and you can
 see all of the different options

0:08:12.140000 --> 0:08:22.380000
 that I have. I'm going to go ahead
 and turn on diagnostics.

0:08:22.380000 --> 0:08:25.780000
 I can archive this to
 a storage account.

0:08:25.780000 --> 0:08:30.960000
 Select the storage account.

0:08:30.960000 --> 0:08:36.980000
 I will keep this in my monitoring
 storage account.

0:08:36.980000 --> 0:08:41.180000
 I can also stream to an event hub
 or stream to log analytics.

0:08:41.180000 --> 0:08:45.500000
 Then I can choose what
 I want to collect.

0:08:45.500000 --> 0:08:49.140000
 You don't necessarily want to collect
 everything because this is taking

0:08:49.140000 --> 0:08:52.700000
 up storage and it is
 costing some money.

0:08:52.700000 --> 0:08:59.500000
 I, in fact, want all
 of these log actions.

0:08:59.500000 --> 0:09:03.180000
 I'm not sure why I went
 in a backwards order.

0:09:03.180000 --> 0:09:10.160000
 I want to store my basic performance
 metrics as well.

0:09:10.160000 --> 0:09:15.800000
 We'll go with 15 days of storage
 for each of these.

0:09:15.800000 --> 0:09:19.160000
 You set it to zero.

0:09:19.160000 --> 0:09:26.300000
 It's going to be unlimited.

0:09:26.300000 --> 0:09:30.800000
 I'm going to save those settings.

0:09:30.800000 --> 0:09:36.280000
 That sets up the diagnostics
 for my server.

0:09:36.280000 --> 0:09:40.820000
 I want to show you one other
 thing for my database.

0:09:40.820000 --> 0:09:43.800000
 I can also set up auditing
 for the database.

0:09:43.800000 --> 0:09:49.540000
 Some of your resources have
 this auditing capability.

0:09:49.540000 --> 0:09:54.540000
 I can set up auditing at the server
 level and it will audit everything

0:09:54.540000 --> 0:09:55.960000
 at the database level.

0:09:55.960000 --> 0:10:00.280000
 Or I can set up auditing for
 an individual database.

0:10:00.280000 --> 0:10:04.080000
 If I'm going to audit a database, I
 can set up where I want that audit

0:10:04.080000 --> 0:10:08.620000
 information to go, storage or
 log analytics or event hub.

0:10:08.620000 --> 0:10:20.180000
 I'm just going to set the subscription
 and set the storage account.

0:10:20.180000 --> 0:10:22.120000
 And the retention.

0:10:22.120000 --> 0:10:28.900000
 Or 365 days of audit information
 on this database.

0:10:28.900000 --> 0:10:32.280000
 And then I can save that.

0:10:32.280000 --> 0:10:38.780000
 In addition to diagnostic data and performance
 data, I also am now storing

0:10:38.780000 --> 0:10:45.760000
 audit data. That's really the story
 with configuring the collection of

0:10:45.760000 --> 0:10:48.560000
 diagnostic data and
 performance metrics.

0:10:48.560000 --> 0:10:54.340000
 Each resource is going to allow you to
 set that up a little bit differently.

0:10:54.340000 --> 0:10:58.340000
 And you need to look at each resource
 and figure out what data you can

0:10:58.340000 --> 0:11:02.540000
 collect, what data you want to collect
 and how long you want to keep that