text stringlengths 0 1.8k |
|---|
[1093.68 --> 1095.26] And this is just like the content stuff. |
[1095.26 --> 1097.26] Then it comes to the CDN. |
[1097.56 --> 1101.00] As I mentioned, the cache status, how many hits versus how many misses. |
[1102.38 --> 1104.72] We can slice and dice by audio requests. |
[1105.60 --> 1116.30] And rather than building dashboards, we can do something even more amazing, which is literally start with a query and keep asking questions and keep getting answers until you understand what's happening. |
[1117.66 --> 1123.00] So this is the first time we've been able to have observability to this level on our CDN. |
[1123.00 --> 1129.96] And so to recap, we leverage it quite well because all requests go through Fastly first prior to hitting our application. |
[1130.46 --> 1136.44] So it would make sense if you make that choice and lean that heavily, trust that much on your CDN. |
[1136.46 --> 1137.24] In this case, we do. |
[1137.24 --> 1137.88] We trust Fastly. |
[1137.94 --> 1140.20] They do an amazing job for us for many years now. |
[1141.48 --> 1147.68] But now we actually have observability into various specifics of how it operates for us where we never had that before. |
[1147.90 --> 1148.34] Correct. |
[1148.74 --> 1153.18] And this is thanks to the details and visibility that Honeycomb gives us. |
[1153.34 --> 1153.68] Correct. |
[1153.68 --> 1154.12] Yeah. |
[1154.12 --> 1154.96] Yeah. |
[1155.14 --> 1159.06] That was one of the big improvements since episode 20. |
[1159.92 --> 1163.26] And we can see the slowest requests. |
[1163.88 --> 1170.08] And we understand that the XML ones, like the sitemap or the feeds, they're the slow ones. |
[1170.62 --> 1173.92] They take five seconds sometimes to load. |
[1173.92 --> 1177.26] And the website is fairly fast. |
[1177.70 --> 1182.24] The only time when it gets slow is when we serve static assets from the website. |
[1182.44 --> 1187.42] So in the Phoenix app, when there's a cache miss in the CDN, it has to go to the app. |
[1187.72 --> 1188.84] Actually, Ingress Nginx. |
[1189.52 --> 1191.12] Ingress Nginx has to go to the app. |
[1191.22 --> 1194.36] And the app has to store like a PNG or JPEG. |
[1194.54 --> 1195.92] It's usually PNG or... |
[1196.48 --> 1197.26] Yeah, PNG. |
[1197.40 --> 1199.46] That's the one that took quite a bit of time. |
[1199.46 --> 1200.66] So I was looking at it. |
[1202.18 --> 1203.14] Was it earlier? |
[1203.80 --> 1204.30] Yes. |
[1204.62 --> 1205.14] Let me find. |
[1205.26 --> 1205.98] It's right here. |
[1207.02 --> 1209.46] It took... |
[1209.46 --> 1211.12] That was an interesting one. |
[1212.30 --> 1213.46] It was... |
[1213.46 --> 1214.82] Icon small. |
[1214.90 --> 1215.84] No, it wasn't that one. |
[1216.10 --> 1216.86] Time elapsed. |
[1216.96 --> 1217.38] This was it. |
[1217.44 --> 1219.54] It was actually a GIF or JIF. |
[1221.06 --> 1225.50] News item, 1.4 minutes to serve it. |
[1226.76 --> 1229.04] That's how long it took to serve that. |
[1229.46 --> 1231.68] News item, GIF or GIF. |
[1232.82 --> 1234.18] All the way to Hong Kong. |
[1235.08 --> 1236.98] So someone from Hong Kong was accessing it. |
[1237.24 --> 1238.28] They were waiting that long, huh? |
[1238.54 --> 1243.72] They were waiting that long because they had to go all the way to our data center in New York. |
[1245.62 --> 1247.82] And I had to wait for it. |
[1247.82 --> 1248.40] It's probably a big... |
[1248.40 --> 1250.82] It's probably a big GIF too. |
[1252.08 --> 1252.32] Yeah. |
[1252.46 --> 1253.26] They always are. |
[1253.42 --> 1256.00] I mean, GIFs are just large files, unfortunately. |
[1256.44 --> 1257.20] At least a meg. |
[1257.32 --> 1258.02] Sometimes 10. |
[1258.72 --> 1259.86] Maybe 50. |
[1260.32 --> 1260.46] But... |
[1260.46 --> 1260.76] Let's see. |
[1260.80 --> 1261.42] How big is it? |
[1261.58 --> 1262.82] We have that as well information. |
[1263.30 --> 1264.72] Body size... |
[1264.72 --> 1267.48] 18 kilobytes? |
[1268.20 --> 1268.54] No. |
[1268.54 --> 1269.36] That's all right. |
[1269.72 --> 1270.08] Megabytes. |
[1270.30 --> 1271.42] So it's like 18 million. |
[1272.18 --> 1272.64] Let's see. |
[1273.70 --> 1275.72] Should we ask Siri to do some math for us again? |
[1276.22 --> 1276.66] Yes, Siri. |
[1277.04 --> 1278.02] 18 million bytes. |
[1278.18 --> 1280.16] They should ask Honeycomb to do that math for us. |
[1280.58 --> 1280.94] Right. |
[1281.04 --> 1282.74] So that's the one thing which you need to set. |
[1283.12 --> 1284.60] So I was setting some... |
[1284.60 --> 1284.80] Yeah. |
[1284.90 --> 1287.46] Some derived queries. |
[1287.86 --> 1288.08] Yeah. |
[1288.16 --> 1288.88] Derived queries. |
[1289.64 --> 1290.44] But... |
[1290.44 --> 1292.10] Let's see. |
[1293.42 --> 1295.00] But not for this specific thing. |
[1296.84 --> 1297.60] 17 kilobytes. |
[1297.68 --> 1298.42] 17 megabytes. |
[1299.34 --> 1299.42] Yeah. |
[1299.50 --> 1302.96] We have a 17 megabyte GIF. |
[1304.22 --> 1306.18] And serving it to Hong Kong, that's how long it takes. |
[1306.18 --> 1306.56] It's pretty heavy. |
[1306.66 --> 1306.82] Yeah. |
[1306.94 --> 1307.14] Yeah. |
[1307.14 --> 1307.62] Yeah. |
[1308.62 --> 1310.54] So the only thing is we do lazy load those. |
[1310.82 --> 1313.56] So you're not actually waiting and user experience. |
[1313.68 --> 1315.50] You can read what the news item is. |
[1315.76 --> 1315.98] And then... |
[1315.98 --> 1316.06] Yeah. |
[1317.16 --> 1320.32] As long as it takes a minute and a half to read, by then, the image is low. |
[1320.42 --> 1321.86] It's still too long, but... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.