Datasets:
willtheorangeguy
/
2023-Ship-It-Transcripts

Dataset card Data Studio Files
xet
 Community
text
stringlengths
0
2.29k
[3832.90 --> 3834.72] But we haven't been able to do that because of this cache layer.
[3834.72 --> 3838.44] Well, our app does run close to our users in the greater Houston area.
[3838.44 --> 3838.74] Yeah.
[3840.16 --> 3840.60] Yeah.
[3840.98 --> 3842.02] It's actually in Virginia.
[3842.58 --> 3843.02] Virginia.
[3843.70 --> 3844.52] Oh, is it?
[3844.76 --> 3845.30] Yeah, yeah, yeah.
[3845.86 --> 3846.96] Well, it shows what I know.
[3847.08 --> 3850.62] It's IAD, the IAD data center, yeah.
[3851.56 --> 3851.88] Yeah.
[3852.54 --> 3856.76] Well, all that to say, like, you know, getting to this direction is challenging.
[3856.88 --> 3862.20] I think the logic in this Postgres sounds fine.
[3862.20 --> 3869.46] I mean, if we were, like you had said, above a larger threshold, you know, a couple megs, not that big of a deal.
[3870.04 --> 3878.58] And if the app's close to the user and there's one, I'm assuming there's probably like one or two primary Postgres writes and then the rest were reads, right?
[3878.62 --> 3882.28] That's how it would set up naturally with Postgres on fly?
[3882.28 --> 3885.18] Yeah, the writes would actually happen on publish.
[3885.68 --> 3890.64] The writes happen on edit, not on first request, which is what happens now with a typical caching.
[3890.78 --> 3893.66] It's like first request, we calculate it once.
[3894.04 --> 3896.42] Now we're not going to calculate it again for 60 seconds.
[3896.52 --> 3897.58] Then we'll calculate it once.
[3898.36 --> 3903.54] This is actually on write is when we're doing the compute, which we've wanted to move to.
[3903.54 --> 3911.20] The other option is to put this on a static file server like S3 and then, you know, manage and blow away different files.
[3911.60 --> 3914.64] But then I started thinking, like, we actually like our URLs how they are.
[3914.76 --> 3918.70] And so then our app would be reading from S3 and responding as a proxy.
[3919.42 --> 3919.52] Yeah.
[3919.84 --> 3921.80] And it's like, well, it's already proxied to Postgres.
[3922.14 --> 3922.64] I don't know.
[3923.28 --> 3928.88] But yeah, we would cache on write versus on read, which makes us have immediate changes.
[3929.00 --> 3931.70] There's no 60 second delay or five minutes or whatever you send it to.
[3932.02 --> 3933.00] And I'm in that camp.
[3933.00 --> 3936.86] I mean, I listen to our show immediately as soon as we ship, you know, the changelog at least.
[3936.98 --> 3940.98] I mean, as just a crazy person, like whenever you ship something, you want to make sure it's in production.
[3941.08 --> 3942.98] The only way to do it is like to test it.
[3943.92 --> 3946.84] And the app I use is Overcast primarily.
[3947.30 --> 3951.44] I don't think I have notifications on because I just hate notifications just generally as.
[3951.98 --> 3952.00] Sure.
[3952.14 --> 3955.14] If I don't have to have notifications on for an application, they're off for sure.
[3955.58 --> 3961.56] But when I do go there, I usually test it on the master feed directly because I listen to master like you should be.
[3961.56 --> 3962.88] Hey, listen, if you're not listening on master.
[3963.00 --> 3963.62] You're wrong.
[3963.86 --> 3965.36] Or plus plus and you'd be even better.
[3965.70 --> 3966.04] Right.
[3966.26 --> 3966.84] This is better.
[3967.34 --> 3970.82] But I'm a master feed subscriber in that regard.
[3971.08 --> 3972.50] And I'm pulled to refresh.
[3972.92 --> 3977.74] And it does take a bit for the new episodes to get there for me, at least.
[3978.06 --> 3982.78] So I'm not like a ship it and 30 seconds, a minute later, it's in Overcast.
[3982.86 --> 3985.04] Like it takes longer than I've counted.
[3985.18 --> 3986.72] Let's just say I haven't actually said they're in Canada.
[3986.84 --> 3987.48] It's like, oh, it's not there.
[3987.60 --> 3988.30] I'll come back later.
[3988.72 --> 3989.58] I come back and it's there.
[3989.58 --> 3996.92] The one thing about this which gets me really excited is that we will double down on PostgreSQL.
[3997.20 --> 3999.16] So we talked about this for a while.
[3999.94 --> 4001.44] Crunchy data is what I'm thinking.
[4001.92 --> 4004.12] But it's not the only way.
[4004.66 --> 4006.24] In what regard are you thinking Crunchy data?
[4006.24 --> 4011.12] I'm thinking a PostgreSQL as a service that scales really, really well.
[4011.46 --> 4013.48] So then the app is all Fly.
[4014.20 --> 4016.16] PostgreSQL is managed via Crunchy data.
[4016.28 --> 4019.42] We have a global presence, nicely replicated, all that nice stuff.
[4019.70 --> 4023.34] And then we consume PostgreSQL as a service at a global scale.
[4023.74 --> 4027.06] Our app runs at a global scale on Fly.
[4027.44 --> 4030.72] And the database is the same, but with someone else.
[4031.08 --> 4034.04] Because the PostgreSQL in Fly, it's not a managed one.
[4034.04 --> 4035.52] It's easy, convenient.
[4035.70 --> 4036.80] We have a lot of advantages.
[4037.28 --> 4040.58] And it's been holding up really well since we set it up.
[4040.80 --> 4041.52] No issues.
[4042.38 --> 4043.66] But we can...
[4043.66 --> 4048.94] I mean, if the app is distributed and if the app gets this level of attention, I think so
[4048.94 --> 4049.78] should our database.
[4050.08 --> 4052.44] Because now these are the two important pieces.
[4052.66 --> 4053.46] We scale the app.
[4053.76 --> 4054.92] We should scale the database.
[4054.92 --> 4060.14] I mean, if, for example, we have all these app instances that connect to the same PostgreSQL
[4060.14 --> 4063.08] instance back in the US, that's not going to be any good.
[4063.88 --> 4064.12] Right?
[4064.16 --> 4067.48] Reading all those megabytes across continents, that's going to be slow.
[4068.62 --> 4071.32] Isn't that the point, though, for like the read servers that are distributed?
[4071.70 --> 4072.66] So we could, yeah.
[4072.68 --> 4076.30] We could add multiple PostgreSQL read replicas in Fly.
[4076.42 --> 4077.24] We could do that.
[4077.42 --> 4077.76] Right.
[4078.86 --> 4079.82] Maybe tune them.
[4080.08 --> 4080.34] Maybe.
[4080.42 --> 4080.82] I don't know.
[4081.16 --> 4081.40] Yeah.
[4081.40 --> 4085.92] Maybe try and understand better what they do, but maybe rather than doing that, we can
[4085.92 --> 4091.22] grow up our approach to databases and go with someone that does this as a service.
[4091.52 --> 4093.72] I know PlanetScale comes up as well.
[4093.80 --> 4094.46] There's like a couple of...
[4094.46 --> 4096.36] We can use PostgreSQL as a service.
[4096.54 --> 4097.50] That's my SQL.
[4097.92 --> 4098.36] PlanetScale.
[4098.48 --> 4100.40] There's one which I know is PostgreSQL.
[4100.68 --> 4102.26] Maybe it's not PlanetScale.
[4102.60 --> 4103.34] What was it?
[4103.96 --> 4104.32] Superbase.