1 00:00:00,180 --> 00:00:06,600 OK, so what we'll do now is we'll create an alert for when an S&P device goes offline, potentially 2 00:00:06,600 --> 00:00:15,120 one of the problems you got with using alerts, hunger Fana, and that is you're unable to create alerts 3 00:00:15,120 --> 00:00:18,020 for four template variables. 4 00:00:18,030 --> 00:00:22,890 So basically what we're going to need to do for everything we want to alert, we're going to need to 5 00:00:22,890 --> 00:00:26,370 create a new graph and add an alert to that graph. 6 00:00:26,370 --> 00:00:31,560 So perhaps in future versions of Gryffindor, we'll be able to use template variables, but not at the 7 00:00:31,560 --> 00:00:34,650 moment, six point four point three. 8 00:00:34,830 --> 00:00:40,500 OK, so let's go down to Dashboard's manage and we'll create a new dashboard. 9 00:00:40,990 --> 00:00:49,290 Let's add a query select measurement SCMP where I can host to Kronenburg. 10 00:00:50,490 --> 00:00:56,810 At the time, we don't care about the visualisations in this one, we're just looking at the numbers, 11 00:00:56,910 --> 00:00:57,990 we're going to create an alert on it. 12 00:00:58,760 --> 00:01:00,390 General, I'm going to call it. 13 00:01:02,810 --> 00:01:03,790 S&P. 14 00:01:06,190 --> 00:01:09,550 Time to coin, Doug. 15 00:01:11,900 --> 00:01:18,090 Alert crit alert is set in the up time to control alert. 16 00:01:18,980 --> 00:01:22,370 I'm going to evaluate for one minute for two minutes. 17 00:01:23,570 --> 00:01:24,950 I'll explain that in the second. 18 00:01:25,190 --> 00:01:29,930 And the query that will run every one minutes will be. 19 00:01:31,260 --> 00:01:33,030 Within the last 30 seconds. 20 00:01:34,550 --> 00:01:35,630 There is no value. 21 00:01:38,390 --> 00:01:45,200 So for every one minute he will run this test, and if that evaluates to true, it will go into a state 22 00:01:45,380 --> 00:01:50,120 and if it's pending for two minutes, it will then go into an alert state and I'll get an email. 23 00:01:50,720 --> 00:01:55,070 So let's just set the email notification channel down here so I get the email. 24 00:01:57,610 --> 00:01:59,920 Save this, I'm going to call it alert. 25 00:02:01,800 --> 00:02:10,110 Monitor safe, excellent, let's add another panel to that very quickly for alert list, OK? 26 00:02:10,800 --> 00:02:14,210 Currently, the status of Milah is unknown. 27 00:02:14,220 --> 00:02:15,480 So I just come. 28 00:02:16,060 --> 00:02:17,490 It'll do a test in a moment. 29 00:02:18,510 --> 00:02:24,210 I'm going to refresh the screen every five seconds and I'm going to change the time range to five minutes. 30 00:02:24,770 --> 00:02:30,180 OK, so these are the last up time values coming through and a test will be run any moment now. 31 00:02:30,180 --> 00:02:31,260 And he should have to to. 32 00:02:31,260 --> 00:02:33,050 OK, there we go. 33 00:02:33,080 --> 00:02:33,540 Okay. 34 00:02:33,540 --> 00:02:34,180 System. 35 00:02:34,320 --> 00:02:35,660 Well, S&P up time. 36 00:02:36,060 --> 00:02:36,630 OK. 37 00:02:38,310 --> 00:02:43,740 Will it sit for a moment, all this fuss, fastforward the video for one minute, so we got some time 38 00:02:43,740 --> 00:02:44,120 there. 39 00:02:49,550 --> 00:02:53,180 AK myosin MP devices just fine, we're getting responses from. 40 00:02:53,250 --> 00:02:54,340 There's no need to alert. 41 00:02:54,800 --> 00:02:58,460 Let's turn off the device now while this off the service. 42 00:02:58,470 --> 00:03:00,170 I'm on my two Koinadugu. 43 00:03:01,080 --> 00:03:04,980 Servoz, I'm going to turn off S&P impeding. 44 00:03:11,470 --> 00:03:20,500 Stop, Pechanga Fana now is a condition is run every minute, so after one minute, if it decides the 45 00:03:20,500 --> 00:03:24,730 last 30 seconds there was no value, it will go into a pending state. 46 00:03:24,820 --> 00:03:27,160 And we'll see that on the graph as well. 47 00:03:30,070 --> 00:03:37,180 Ago, within the last 30 seconds, there was no value, so it's impending, it may to be impending for 48 00:03:37,180 --> 00:03:39,340 two minutes before we get another. 49 00:03:46,480 --> 00:03:53,860 Like I said, that's been pending for two minutes, we're now in the alert phase alerting this to my 50 00:03:53,860 --> 00:03:54,430 emails. 51 00:03:55,740 --> 00:04:03,150 Let we go to my spam box, so I haven't quite listed it yet, incident pay up, time to Koinadugu alert. 52 00:04:05,570 --> 00:04:09,650 And we get was no data, I get a little graph as well. 53 00:04:10,000 --> 00:04:10,750 We're interested. 54 00:04:12,350 --> 00:04:14,370 Excellent, later. 55 00:04:15,130 --> 00:04:18,350 Let's now turn S&P back on. 56 00:04:23,450 --> 00:04:31,520 Go back and see what happens next, one minute check if the last 30 seconds were OK, you will go back 57 00:04:31,520 --> 00:04:34,940 to OK, OK, OK, that's a little green dot. 58 00:04:35,440 --> 00:04:36,740 I check my emails again. 59 00:04:37,940 --> 00:04:41,170 Here we go, alerting, OK? 60 00:04:42,760 --> 00:04:46,630 OK, it's an impact on to control, and we got a graph. 61 00:04:47,410 --> 00:04:49,120 Being great again, excellent. 62 00:04:50,250 --> 00:04:52,380 Let's look at those settings again. 63 00:04:53,920 --> 00:04:54,830 More specifically. 64 00:04:54,850 --> 00:05:04,900 OK, so this is the alert rule evaluates every one minute if this evaluates to true condition. 65 00:05:06,210 --> 00:05:10,420 It goes on, depending if it's been pending for two minutes, it will go into alert. 66 00:05:10,680 --> 00:05:17,310 Now we can either sit this at zero and that will go into alert as soon as it goes, depending. 67 00:05:17,850 --> 00:05:23,580 Or we can just leave that even to one minutes or even or even ten minutes. 68 00:05:23,580 --> 00:05:24,300 It's up to you. 69 00:05:24,540 --> 00:05:29,790 If you have these values too small, you you'll get what's known as alert flapping. 70 00:05:31,690 --> 00:05:41,770 It's error, OK, error, OK, what often and since S&P actually runs on UDP by P protocol, it's an 71 00:05:41,770 --> 00:05:48,620 unreliable communications protocol, so it'll be moments when there is no response for S&P. 72 00:05:49,390 --> 00:05:54,460 So, I mean, this condition now is likely to happen quite often. 73 00:05:54,460 --> 00:05:58,930 So, you know, I can also sit there for one minute if I wanted to. 74 00:05:59,480 --> 00:06:01,870 So it's kind of best to have some buffer there. 75 00:06:02,260 --> 00:06:06,220 So I'm going to have one one one and save that. 76 00:06:09,860 --> 00:06:15,260 I'm going to create another two conditions for my resident pay devices, remembering we have to do this 77 00:06:15,260 --> 00:06:20,540 manually in Gravano because we're unable to use template variables. 78 00:06:20,660 --> 00:06:22,190 I was going to fast forward that bit. 79 00:06:31,820 --> 00:06:38,200 The guy there is three separate S&P up time alerts credit. 80 00:06:39,440 --> 00:06:47,120 Quite apart from megafauna server as well, I call on 27 zero zero one, I mean, it should probably. 81 00:06:48,140 --> 00:06:56,020 Test your grandfather, Serbo availability, using a different system, but doing it here anyway. 82 00:06:57,090 --> 00:07:03,530 Very good, so we can see all these things are running and I'll get alerts when I need to.