Call —
Transcript streaming
0:00 Caller: Hello, Jed.
0:00 You: This call will be recorded.
0:05 You: Hello.
0:08 You: Yes.
0:10 Caller: Hey, Jed, sorry for the double dial, I think I tried to connect with you a second ago and I got disconnected, but my name is Jared Knox. I'm calling from Timeskills.
0:20 Caller: db how are you doing today um so i'm reaching out to you jett i understood that you're one of
0:20 You: Hello, Jared. How can I help you?
0:28 Caller: the senior research engineers over there at row technologies is that so correct
0:35 Caller: and i was looking into your lincoln profile a little bit you kind of seemed like you might
0:38 Caller: have been the appropriate
0:40 Caller: person to kind of reach out to about this but I'm not sure if you're familiar with
0:40 You: I'm
0:44 Caller: timescale DB but it's the database that's built on top of Postgres you just
0:49 Caller: extended to handle timestamp or time series specific workflows so you're not
0:53 Caller: having to split relational and analytic data between two different systems
0:56 Caller: since it is built on top of Postgres but in
1:00 Caller: feel free to correct me if I'm wrong, but it seems like road technologies might be collecting
1:00 You: We're going to be.
1:02 You: I'm
1:03 Caller: a amount of time series or timestamp data. But typically as, you know, that those time series
1:12 Caller: data is grow within the organization, query times start to get a little bit longer, a little bit more
1:17 Caller: taxing. And time...
1:20 Caller: Time scale is just ensuring that those are remaining as fast as possible without leaving Postgres, since it is an extension.
1:20 You: Thank you.
1:25 Caller: But based off what I knew about your title and the organization, I felt like you might have just been a good individual to reach out to, maybe get some time on the calendar for next week and dive a little bit deeper.
1:35 Caller: But is that something that you'd be open to potentially doing?
1:40 Caller: RK. Okay. And does that mean that you guys are on two different database platforms at the moment, or is it just going to be on primarily on that own or that individual platform?
1:40 You: not really we do predominantly parquet for our time series work
2:00 Caller: Mm-hmm.
2:00 You: Data lives in a Postgres instance. So it needs the read, update, and delete portions of crud. The analytics instance is predominantly parquet and then a series of query runners that can access it. So Ducti B, Trino, Presto, Click House, Star Rocks.
2:10 Caller: Mm-hmm.
2:13 Caller: Perfect.
2:16 Caller: Thank you.
2:18 Caller: Thank you.
2:20 Caller: Okay, if you could do all of those, if you could do both your relational and your analytic data without having to leave Postgres and kind of eliminate that.
2:20 You: whatever fits the specific need, but most importantly, we can very quickly migrate off a vendor when the need comes.
2:40 Caller: extra system, do you think that that would be beneficial to the organization?
2:40 You: Postgres doesn't really scale in our environment beyond three or four terabytes.
2:50 You: Yes, you can scale it vertically, a bit put bigger and heavier machines, but there's still a limit.
3:00 Caller: exactly what we're doing at timescale.
3:00 You: We're going to be.
3:02 Caller: Postgres is not built to handle time series
3:06 Caller: workflows originally, but at timescale,
3:10 Caller: it's an extension that makes it basically geared and suited
3:14 Caller: to do what you're doing on those additional systems
3:17 Caller: without having to leave Postgres.
3:18 Caller: So imagine doing everything you're doing.
3:20 Caller: on those other systems without having to leave Postgres and now you're doing everything on one platform.
3:20 You: So what exactly do you think Rowe's data shape looks like to justify time scale over
3:26 Caller: That's essentially what we're doing over here.
3:40 Caller: Well, that would be something that I would love to learn a little bit more about from you.
3:40 You: the existing solutions.
3:50 Caller: I think I could use a bit of...
3:50 You: Something about your research suggested that Roe could benefit from time scale.
3:56 You: I'm trying to understand what those assumptions were.
4:00 Caller: Just the, I suppose, your guys' platform, what you guys are doing online, what I can see from your website, kind of rang a bell in my head, made it appear as if there was some sort of data or time series collection being done.
4:00 You: you're not wrong but as a quick reminder
4:20 Caller: Right.
4:20 You: business of handling customers money so misplacing a penny is unforgivable which means separation of concerns becomes especially important and so the application database the ledger things that touch money are very carefully engineered to ensure that the
4:40 Caller: at any point whatsoever is that is that correct okay okay okay
4:40 You: record of whose money is where becomes immutable. Shifting that to another database platform,
4:47 You: even if it's just a plugin, is not something we would do. Correct, because we are beholden to
4:58 You: regulators.
5:00 Caller: So, if I'm understanding correctly, the data is already on Postgres.
5:00 You: The additional system is the extension.
5:10 Caller: So it wouldn't have to, you wouldn't necessarily be adding an additional system, correct?
5:14 Caller: Even if it's as an extension, it was still right.
5:17 You: That's the part that we would need to get
5:20 Caller: Understood. Okay. Okay. Well, in that case, it sounds like I suppose it might not be the best use case from what you're explaining to me. So completely understood. What I can do is I'll go ahead and send you over some additional.
5:20 You: certified for.
5:40 Caller: information just in case so you do have that in your back pocket I have your
5:40 You: Thank you.
5:42 Caller: email and then if you find for whatever reason it is interesting at that point
5:48 Caller: you can reach reach out to me but I appreciate you kind of me still it kind of
5:53 Caller: filling me in on what the situation was because that that is very very helpful but
5:58 Caller: if that's the case I can go ahead
6:00 Caller: let you go jed but i do appreciate you taking the brief phone call and i hope you enjoy the rest
6:00 You: You as well.
6:04 Caller: your day all right bye bye
6:06 You: Take care.