Meeting Title: DBT Model Deprecation Sync Date: 2025-07-02 Meeting participants: Demilade Agboola, Caio Velasco


WEBVTT

1 00:00:29.350 00:00:30.909 Caio Velasco: Hey? There, Mandy! How’s it going.

2 00:00:34.394 00:00:41.860 Demilade Agboola: Hey, it’s going. Okay, just a lot of meetings. I think I’ve had like, almost like 3 h worth of meetings. Yeah.

3 00:00:42.060 00:00:42.670 Caio Velasco: Oh, my God!

4 00:00:43.010 00:00:43.740 Demilade Agboola: Yeah.

5 00:00:44.180 00:00:45.499 Caio Velasco: Hate those things.

6 00:00:46.282 00:00:51.447 Demilade Agboola: Yeah, at this point my days are like that. Most of my days are like that.

7 00:00:52.070 00:00:55.508 Caio Velasco: Okay, let’s be. Let’s be quick.

8 00:00:55.980 00:00:56.660 Demilade Agboola: Yeah, sure.

9 00:00:58.200 00:01:03.850 Demilade Agboola: So for this, what I’m what I’m trying to say is, if

10 00:01:07.660 00:01:14.120 Demilade Agboola: if you’re able to communicate like what appears to be wrong, or appears to be like

11 00:01:17.050 00:01:18.699 Demilade Agboola: what appears to be

12 00:01:19.120 00:01:29.760 Demilade Agboola: the disparity, like, what’s the cost of the disparity? It makes it easier for me to say, oh, just like use manifesto Json, or just use like the Dvt audit

13 00:01:30.653 00:01:36.290 Demilade Agboola: but like as it is right now, I’m not exactly sure what like what is causing that difference.

14 00:01:37.570 00:01:38.740 Caio Velasco: No problem.

15 00:01:39.030 00:01:46.917 Demilade Agboola: Yeah. So it makes it hard for me to, unless I’m going to have to dive in and start going through it like bit by bit, and then say everything.

16 00:01:51.610 00:01:54.190 Caio Velasco: Let me share the screen. Then.

17 00:01:55.050 00:02:02.160 Caio Velasco: okay, I can guide you with briefly, and then we can see. And you, with your experience, can see like what would be the best thing to do.

18 00:02:03.350 00:02:04.840 Caio Velasco: Do. Do you see my screen.

19 00:02:05.190 00:02:06.019 Demilade Agboola: Yeah, proceed.

20 00:02:06.500 00:02:08.940 Caio Velasco: Yeah, let me put it larger.

21 00:02:11.270 00:02:13.380 Caio Velasco: Okay, so

22 00:02:13.650 00:02:22.950 Caio Velasco: yeah, so we have this main sheet now, which has been updated last week or so because Utm asked me to

23 00:02:23.070 00:02:27.569 Caio Velasco: do the usage, which is basically all this

24 00:02:28.272 00:02:36.679 Caio Velasco: based on 180 days that things were queried or scanned because Redshift used the the scan

25 00:02:36.820 00:02:37.940 Caio Velasco: language.

26 00:02:39.170 00:02:48.300 Caio Velasco: Okay, so yeah, so this is the 1st thing we have. Some are true, a lot of false. This would be the 1st thing we deprecate which would be my my next task.

27 00:02:48.540 00:02:51.409 Caio Velasco: Then I would be basically deprecating.

28 00:02:51.790 00:02:52.990 Caio Velasco: Well.

29 00:02:53.660 00:03:00.370 Caio Velasco: all these things, and then I don’t think we care about this part. So this is my 1st question, does it make sense? Because we do have

30 00:03:00.510 00:03:02.569 Caio Velasco: things from analytics here?

31 00:03:02.810 00:03:04.859 Caio Velasco: But they haven’t been queried.

32 00:03:05.200 00:03:09.860 Caio Velasco: So I am assuming that we can archive them.

33 00:03:12.250 00:03:19.770 Demilade Agboola: And I think we should make exceptional things that analytics, because things in analytics are being created by Dbt

34 00:03:20.550 00:03:27.710 Demilade Agboola: even though they’ve not been used in a while not been query in a while. We need to, I think, for those ones. We might need to ask questions as to

35 00:03:28.100 00:03:30.140 Demilade Agboola: why that is.

36 00:03:31.900 00:03:37.920 Caio Velasco: Okay. As, for example, if I were to just put these ones, we have

37 00:03:38.650 00:03:43.410 Caio Velasco: 230 like, it’s a lot. And like all views and tables.

38 00:03:43.700 00:03:48.060 Demilade Agboola: Yeah, that’s a lot actually, is a lot.

39 00:03:51.160 00:03:59.650 Caio Velasco: Because this came basically. And this there’s no space, I believe to be mistaken, because this came directly

40 00:03:59.840 00:04:02.270 Caio Velasco: from redshift, from querying.

41 00:04:02.780 00:04:12.549 Caio Velasco: The let’s say the the the analogous of information. Schema in redshift.

42 00:04:15.360 00:04:16.140 Caio Velasco: So.

43 00:04:16.870 00:04:21.410 Demilade Agboola: Give me one second. Let’s see if I can see one of these tables in like the Dvc model.

44 00:04:22.840 00:04:25.850 Demilade Agboola: If it still exists as a model.

45 00:04:26.120 00:04:27.679 Caio Velasco: Oh, that’s fine.

46 00:04:27.680 00:04:30.280 Caio Velasco: I do that. Okay. Otherwise I can open here.

47 00:04:31.670 00:04:39.390 Demilade Agboola: I’m actually just looking at the the easy post underscore base.

48 00:04:40.750 00:04:43.069 Demilade Agboola: All right. So my guess is okay.

49 00:04:44.010 00:04:46.590 Demilade Agboola: These are models that have been renamed.

50 00:04:47.520 00:04:53.349 Demilade Agboola: So the old models that exist. But like they’re not running based on the score convey.

51 00:04:55.200 00:04:59.130 Demilade Agboola: Yeah, okay, yeah, we can delete them. We can delete them.

52 00:04:59.790 00:05:02.099 Caio Velasco: Okay? And are you sure for all of this.

53 00:05:03.740 00:05:07.069 Demilade Agboola: Let me see, because we have even like.

54 00:05:07.070 00:05:10.290 Caio Velasco: When message stuff orders.

55 00:05:11.760 00:05:14.330 Demilade Agboola: Let me see. Oms, orders or Ms.

56 00:05:14.630 00:05:15.840 Demilade Agboola: Oh, there it is.

57 00:05:18.180 00:05:21.749 Demilade Agboola: Yeah, there’s no, there isn’t an oms orders anymore.

58 00:05:23.780 00:05:28.070 Demilade Agboola: Let me see, where am I? Stop order.

59 00:05:30.140 00:05:36.190 Demilade Agboola: I can look as there isn’t almost a lot of items guessing. It’s just not running

60 00:05:47.220 00:05:48.350 Demilade Agboola: interesting.

61 00:05:50.860 00:05:58.580 Demilade Agboola: But we have a list of Dbt models. Is it possible to just like search against this and remove them.

62 00:06:02.170 00:06:03.449 Caio Velasco: Sorry I didn’t get it.

63 00:06:03.850 00:06:04.929 Demilade Agboola: Only such

64 00:06:06.270 00:06:12.199 Demilade Agboola: for the one stop here like it’s like the manifest Jason, which is part of what I was speaking to.

65 00:06:12.390 00:06:17.119 Demilade Agboola: but like they appear in the Dbt. Models.

66 00:06:21.310 00:06:27.450 Caio Velasco: Well, if we just look at the manifest, then well, with the S.

67 00:06:27.760 00:06:33.900 Caio Velasco: We still have a lot, I mean, not a lot, but a few like 62 that

68 00:06:34.390 00:06:36.749 Caio Velasco: are not being queried in redshift.

69 00:06:37.770 00:06:40.230 Caio Velasco: but it is present in the manifest.

70 00:06:41.880 00:06:46.320 Demilade Agboola: Also those rest of your queries also count like

71 00:06:46.900 00:06:53.149 Demilade Agboola: if they’re used by looker, or if they’re being used by any other like, if they even use within Dbt. Upstream.

72 00:06:54.870 00:06:56.320 Caio Velasco: Let me check. Here.

73 00:06:59.910 00:07:03.240 Demilade Agboola: But like, for instance, if something’s a downstream model in Dvt.

74 00:07:03.570 00:07:09.920 Demilade Agboola: so an option model in Dvt and downstream models use it good.

75 00:07:09.920 00:07:11.789 Caio Velasco: From the queried pipe.

76 00:07:12.120 00:07:18.630 Caio Velasco: Probably I’m not sure, because I don’t know much about this, but we’ll have a lot of users here. Look a user.

77 00:07:18.750 00:07:20.910 Caio Velasco: a polyatomic axis.

78 00:07:25.110 00:07:26.540 Demilade Agboola: Can I go? Can you go?

79 00:07:27.940 00:07:31.740 Demilade Agboola: So yeah. Dbt, part as well.

80 00:07:33.400 00:07:35.190 Caio Velasco: Yeah, we brought the stage.

81 00:07:40.780 00:07:42.610 Caio Velasco: Okay, mode analytics.

82 00:07:44.110 00:07:45.580 Caio Velasco: This is probably good.

83 00:07:45.960 00:07:48.579 Caio Velasco: Read replica. I don’t know. What is this.

84 00:07:54.950 00:07:58.360 Demilade Agboola: Okay, alright. So what we’re gonna do is I think we should

85 00:08:00.740 00:08:11.860 Demilade Agboola: the ones that are not being used by anyone’s. So it’s not like it’s not true, for is in manifest is not true, for is used by Dvt. Let’s get rid of those ones. Those ones we can be

86 00:08:12.240 00:08:19.930 Demilade Agboola: like those ones that were like, oh, it’s not being used, and it’s let me see.

87 00:08:20.780 00:08:24.779 Caio Velasco: Yeah. And that’s where we started. If we would put this as precedence here.

88 00:08:24.920 00:08:30.039 Caio Velasco: because theoretically it just is scripted and and then join here.

89 00:08:34.110 00:08:42.340 Caio Velasco: Then it would be putting all this out. But then you from yours. There is some yes, which I also don’t know why. There was differences.

90 00:08:43.620 00:08:47.359 Demilade Agboola: Yeah. So that’s why I was saying like it would be. Also, if you like.

91 00:08:47.520 00:08:53.990 Demilade Agboola: Look into them like, is anything of like those models have their renamed now, because

92 00:08:54.970 00:08:59.319 Demilade Agboola: that’s also protect. That’s also a possibility. So like, if you can look through like

93 00:08:59.710 00:09:04.939 Demilade Agboola: Dbt’s the repo right now, and see if you can find the names of these things.

94 00:09:05.360 00:09:15.170 Demilade Agboola: if you can’t find them, then maybe that’s what’s happened. Maybe like, Oh, these number, these models don’t exist anymore. But they’ve been renamed because Emily does love things. She’s.

95 00:09:22.050 00:09:24.310 Caio Velasco: Yeah, apparently this.

96 00:09:24.650 00:09:26.000 Caio Velasco: It’s not there.

97 00:09:28.400 00:09:31.129 Demilade Agboola: That shouldn’t be there, because it’s it’s even N.

98 00:09:31.640 00:09:33.390 Caio Velasco: Yeah, this is also my meeting. Right?

99 00:09:34.780 00:09:35.370 Caio Velasco: Alright.

100 00:09:35.370 00:09:36.249 Demilade Agboola: And as well.

101 00:09:36.250 00:09:38.100 Caio Velasco: Yeah, you you’re looking for this one. Right?

102 00:09:38.550 00:09:39.095 Demilade Agboola: Here

103 00:09:40.000 00:09:42.500 Caio Velasco: A good one here, Fedex.

104 00:09:51.290 00:09:58.039 Caio Velasco: not in the main name, just as a city name.

105 00:09:58.610 00:09:59.170 Caio Velasco: Hmm!

106 00:09:59.170 00:10:04.880 Demilade Agboola: Yeah. So that that’s why I’m guessing. Some of these things are like, maybe things where the names of these things have changed.

107 00:10:05.000 00:10:08.119 Demilade Agboola: That’s why, I said, like getting an idea of what’s going on.

108 00:10:08.370 00:10:14.210 Demilade Agboola: So the main thing of these things have changed. And that’s why they don’t appear anymore.

109 00:10:15.200 00:10:16.070 Caio Velasco: Okay.

110 00:10:16.570 00:10:17.280 Demilade Agboola: Yeah.

111 00:10:17.790 00:10:22.389 Demilade Agboola: So obviously, if the names have changed, then, yeah, it doesn’t matter anymore. Because,

112 00:10:25.560 00:10:31.190 Demilade Agboola: if the names have changed doesn’t matter anymore. Because, like Dbt will not produce those models anymore. So that’s fine.

113 00:10:32.660 00:10:38.580 Caio Velasco: True, true. So theoretically, what we are saying is that we are disregarding this one.

114 00:10:39.270 00:10:40.040 Demilade Agboola: Yes.

115 00:10:41.290 00:10:44.029 Caio Velasco: Okay, okay, okay, you wanna put this red.

116 00:10:44.500 00:10:52.409 Caio Velasco: Okay? So we keep this this. So basically, our 1st layer of deprecation will be

117 00:10:52.530 00:10:59.969 Caio Velasco: not only the false here, but the ones that are not presently manifest. So basically, this 2 filters.

118 00:11:01.120 00:11:01.810 Demilade Agboola: Yeah.

119 00:11:02.070 00:11:11.649 Caio Velasco: Okay, and then for well, we start with this one. And then we see about the accurate models.

120 00:11:14.220 00:11:21.819 Caio Velasco: okay, I still have to update this. So this is, don’t worry about this one now. But then we can move to then second layer and include the accuracy idea.

121 00:11:23.150 00:11:25.860 Caio Velasco: Yeah, okay, I think with this, I’m good.

122 00:11:26.020 00:11:28.210 Caio Velasco: Do we need anything else? I don’t think so. Right.

123 00:11:28.622 00:11:30.270 Demilade Agboola: No, no, that’s fine.

124 00:11:30.270 00:11:34.501 Caio Velasco: Okay, cool. So, man, have a good day, and I hope you don’t have any meetings again.

125 00:11:35.080 00:11:38.050 Demilade Agboola: Have to get some models out, so we’ll see.

126 00:11:38.050 00:11:39.450 Caio Velasco: Okay. Okay. Cool.

127 00:11:39.450 00:11:39.890 Demilade Agboola: Alright. Thank you.

128 00:11:40.050 00:11:40.569 Caio Velasco: See you.

129 00:11:41.420 00:11:41.980 Caio Velasco: Good day.