Technical Troubleshooting and Optimization Discussion

0:50:37

The video captures a detailed discussion among a team focusing on resolving slow website performance, investigating errors, and optimizing WordPress and HTML landing pages. The team deliberates on potential solutions, examines data from various tools, and plans steps forward to solve performance issues while discussing refunds for monitoring charges.

Technical Troubleshooting and Optimization Discussion

Timeline




	00:00	(Beginning of video)

	00:00	Introduction and Initial Problem Report

	00:00	From that though, I mean I get report.

	00:03	Well like a report link every now and then.

	00:06	But you know, I, I haven't, I haven't received anything from there.

	00:10	I, I sent them the, the email on the 13th. Yeah.

	00:14	Discussion on Handling Credits and Refunds

	00:14	With the information, you know, that you, you sent me. Yeah.

	00:19	But yeah, no answer from them.

	00:22	Well, what they told me was they just couldn't talk to me, but they said that's fine.

	00:26	You know, it happens sometimes.

	00:27	We apologize.

	00:29	Thank you for bringing that to your attention.

	00:31	We'll, you know, we'll definitely work with you guys to give you a credit. Yeah.

	00:36	So I said, well, is it a credit or a refund? So.

	00:40	No, no, we'll refund the, the amount minus what, what you actually, you know, we'll calculate basically what you owe. Yeah.

	01:46	Technical Issues and Slowness

	01:46	You know, slowness. Right.

	02:21	Storage.

	02:49	Chris, it's not the switch mechanics.

	04:04	Server and Network Analysis

	04:04	AWS West.

	04:05	The AWS East.

	04:08	Italian, GCP East. Right.

	04:11	So if I take California and, and west.

	04:20	Right. As you can see, it's always been green. Right?

	04:22	It's always green.

	04:26	Now let me show you something though.

	04:28	Hey, why can I hold on a second.

	04:30	I can't interact with this.

	04:34	All right. Can you guys still see it?

	04:37	Yeah, yeah, I can see it.

	05:29	Storage and Server Mechanics

	05:29	Okay, Saki triggered. Let's see what this is.

	05:35	This is like 7:38pm There was.

	06:05	The website did not load in less than 120 seconds.

	06:09	Okay.

	06:47	Just a home page.

	07:20	Website Loading Times and Error Identification

	07:20	Are you saying that is it loading also like WordPress and other things or just HTML?

	07:27	I think it's both.

	07:30	It's both. Yeah.

	07:31	Because you know, they have been presenting slowness in both of the, you know, HTML and also in the WordPress server.

	07:41	So this is not a good test then.

	07:43	So I think, I think if Felipe, if you could send me a direct page that is just HTML like not running on the WordPress server.

	07:54	Yeah, I can send them to you.

	08:08	Measurement rendering on WordPress.

	08:14	Look at application stack.

	09:49	HTML vs WordPress Performance

	10:28	Yeah.

	10:45	Modules backup service line 100 and it says basically NS error failures.

	10:52	So for example C WordPress.

	11:03	Backup Service Error Investigation

	11:03	Yeah, I don't know.

	11:04	I mean in that case I just shared with you one of the pages that is an HTML page on the chat.

	11:12	That one is supposed to just be HTML static, not touching the WordPress server.

	11:20	In regards to that one that says the backup service.

	11:24	That's something that probably we need to speak with Benjamin because let me do that then.

	11:31	Let me, let me copy and paste these into an email premier.

	11:59	So to. To Felipe and then, and then.

	12:14	Yeah, then we'll keep this one.

	12:25	Those are events, right?

	12:27	Yep, exactly. Those are all events.

	12:29	So, like, let's go back.

	12:31	Let's go back to.

	12:44	Okay, two errors.

	12:48	Now let's see. What is.

	13:03	That was a picture and console error?

	13:11	No, because I guess one of the things when it was doing the get.

	13:15	It was just rendering the whole page.

	13:19	The page is trying to render.

	13:21	So it's trying to, like, bring up all the.

	13:23	All the images.

	13:24	And look, web server is down.

	13:27	But it was because of this component failure.

	13:29	Backup service sys. Mjs.

	13:36	Yeah, this would. This.

	13:38	I guess this would be a Benji question. Let me.

	13:40	Let me try to add him. He's. He's right here. Hold on. Okay.

	14:08	Engagement with Team Member Benji

	14:08	Monitors. Yeah.

	14:10	Monitor list.

	14:16	Synthetics.

	14:19	Okay, so Benji's here.

	14:21	Hey, Benji. Hey, Benji. Hello.

	14:26	Hi. Welcome, Benji. Welcome, Benji. Hi. How are you? Good, man.

	14:31	I don't know if you can see my screen, but we're looking at the datadog measurements and so one of the things.

	14:44	Which is just an HTML. HTML?

	14:56	Yeah, yeah.

	15:11	I don't know. I don't know.

	15:17	I'm sorry, can you mute your mic?

	15:37	So what I was saying, Benji, is that now I have.

	15:42	Felipe gave me basically a different URL to hit.

	15:45	So we'll create a separate measurement that doesn't involve WordPress. It doesn't involve.

	15:50	It's just basic HTML page.

	15:52	Because as you can see, all these little red dots, those are failures.

	15:58	This is when basically the threshold is met that the page didn't render in a certain amount of time.

	16:05	Monitoring and Error Tracking

	16:05	So what we're doing is we're going through each of these and clicking on them and trying to get to the source of the actual alert.

	16:15	And so if we come here, this is from the west coast, but it fell on the east as well.

	16:24	And then if we go here, look at the error.

	16:29	It's this same module right here.

	16:38	We were asking us, do you know what that module is?

	16:40	Like, what function this.

	16:42	This is providing?

	16:46	I have no idea about this.

	16:48	I'm actually looking at this message for the first time.

	16:54	Okay.

	17:00	If you see.

	17:02	Is the same. Is the same.

	17:06	Yeah. Yeah. It seems like every one of them.

	17:13	Every one of the.

	17:15	So if we go back even two months.

	17:20	Let's go. Let's go.

	17:25	And let's go all the way to the.

	17:42	There we go. Okay.

	17:49	Let me see something, though.

	17:53	Is it always important?

	17:58	Interesting. Wait, is there any way to just like, you know, check for different.

	18:08	Wanted to make sure it wasn't always from Oregon, you know.

	18:11	Now, if it's always from the west coast, though, it could be that Los Angeles.

	18:19	Geographical Impact on Performance

	18:19	It could be that some traffic conditions, you know, network conditions from The west coast coming to.

	18:27	To this Miami data center caused it.

	18:32	Yeah, look, it's always the west coast, see, which I'm not saying, I'm not saying that it's necessarily that.

	18:46	But obviously there's, there's going to be higher latency for traffic West Coast.

	18:52	So there is obviously something going on on the application, you know, in the environment that is not apparent to people on the East Coast. Right.

	19:01	Because it's not crossing the threshold and it's more seen from people on the west coast because there's a longer distance for that package of travel.

	19:10	Okay, here's Virginia. That's good.

	19:13	So let's see what Virginia triggered.

	19:19	Yeah. So basically what are we trying to achieve by doing this?

	19:24	So we're trying to.

	19:25	Well, first of all, we're trying to establish that whenever we're seeing slowness, we can quickly eliminate the network.

	19:34	Because whenever you're having issues, if you're loading up, you're going against a web server and you're loading that static page on that web server that's hosted up issues. It's not.

	19:52	But if you're going against another measurement that is actually loading WordPress for instance, and that's the magnet that is now, you know, causing showing the slowness, then we know it's the WordPress application.

	20:02	Recent Website Migration and its Effect

	20:14	Would you agree or disagree?

	20:19	Okay.

	20:26	It is actually not on.

	20:27	Net anymore.

	20:30	The whole AI Edu is not.

	20:33	Yeah, this website is moved to the. To the new hosting.

	20:36	It's not on the atomics.

	20:40	Okay. Okay. I don't want.

	20:41	I don't communicate with that. So.

	20:43	So you're saying the entire somewhere else.

	20:48	Yeah, absolutely. Okay.

	20:50	In the same.

	20:51	Like behind switches. No, no, no. It's.

	20:55	It's offshore data center.

	20:57	Not offshore, but yeah, it's in New York.

	21:00	Okay. Okay, so do we have URLs are still sitting on.

	21:07	Yeah, yeah. You can.

	21:08	You can use it WP AU Edu. That.

	21:12	That's live also you can use it to test the version that is sitting on Danics production.

	21:20	Are there any production?

	21:23	No, no, no.

	21:27	Maybe you can text with the.

	21:29	The aspx, but that has to do with the.

	21:32	That is not the main website.

	21:34	I think that has never been reported.

	21:36	You know, a problem also that much.

	21:41	They are still on Nutanix.

	21:43	Okay, so let me ask a different question.

	21:47	Are you guys still having performance issues?

	21:58	I'm not sure on this one. But when.

	22:00	The last time when we saw issue related to performance, it was on the MySQL front and then you also stepped in and we had reverted the kernel of Linux back to an older kernel and that's how we resolved that.

	22:15	And before that, it was an issue that was at the disk iops. Okay.

	22:21	But after that, you know, we did not troubleshoot that much on that machine.

	22:26	And honestly, I'll tell you something.

	22:28	If I try to transfer the database, a small three terabytes.

	22:33	Sorry, three gigabytes database from Nutanix VM to another data center, okay.

	22:39	It is four to five hours of work.

	22:41	It is that slow.

	22:43	If I try to copy files, it moves it in mbs like 1 or 2, 3 mbs per second.

	22:48	I've completely given hope on that product.

	22:51	Honestly, that's how the performance was.

	22:55	It's a nightmare to just copy the data off from that machine onto the Internet.

	22:59	I've used rsync. Okay.

	23:01	I've used SCP copy, I've used normal FTP copy.

	23:07	All of them are still.

	23:09	Okay. Kevin, one question.

	23:11	Since when did it move?

	23:16	When we moved like four days ago.

	23:22	Four days ago.

	23:23	Four days ago.

	23:26	Okay, so this data is just moved the main website and the MySQL server, it is hosting WordPress.

	23:35	Okay, but this data is still valid.

	23:37	We're not looking at four days go.

	23:41	Performance Tools and Refund Process

	23:44	No, the URL is not under dings anymore.

	23:48	No, I understand that, but you just said it was moved four days ago. Days ago. Four days ago. Yeah. Right, right.

	23:53	The data we're looking at.

	23:56	Oh, then you're looking at the. The correct.

	23:59	Which you're trying to figure out.

	24:00	Right, Right, right.

	24:03	Okay. Oh, and we don't know what this is though, right? All right.

	24:11	I have no idea on this one question.

	24:32	The money or credit.

	25:02	Okay, so should we.

	25:10	Should we call them instead of like just send the email?

	25:15	Say again? Should we just call them instead of sending the email?

	25:18	I think so, yeah. Yeah.

	25:20	I mean, it's been what, it's been about three days, right?

	25:23	Yeah, we send that the 13.

	25:25	Yeah, I would call.

	25:28	Hey guys, I need to step away for the dinner.

	25:33	Is there something.

	25:35	No, no, we're good. We're good.

	25:36	Yeah, we just wanted to know if you knew about this. This error code.

	25:40	But you don't know. Yeah, no.

	25:42	No idea about these ones. Okay. All right.

	25:46	Okay, thank you. Engine. Yeah, thank you. Let me know. Okay, bye. Bye.

	26:45	Then we were doing it against Ultra and.

	26:50	And it was fine.

	26:51	So we learned that it was like the network stack on that one vm.

	26:56	You know, we were getting like normal speed, so.

	26:59	No, yeah, it's fine.

	27:01	I mean, if it's moved off, it's a move point, but.

	27:05	Oh, yeah, see, laprema Seeing it don't.

	27:17	It's a software.

	27:18	Right, Exactly.

	27:46	Yeah. I mean, and the error is still the same console error backup service js.

	28:16	I want to see what that is.

	28:17	Let me see what this model is for.

	28:22	Required return failure code.

	28:29	I'm just going to send this to you guys in an email.

	28:42	I wonder if it's also like a browser issue, you know?

	28:49	Yeah, I was checking and some, some places it says it's a Firefox issue.

	28:56	Now let me see if it says the message.

	29:01	NSerror failure is a generic error in Mozilla's code base.

	29:07	Okay.

	29:10	So I wonder if the people who are having slowness issues.

	29:16	I wonder if it falls down to that.

	29:18	Like, I wonder if it's ever been reported.

	29:20	You know, we used to have, when I worked at Omnation, we used to have dealerships call us all the time, call our help desk saying, oh, it's slow, it's slow.

	29:29	And then we started asking, telling the help desk to ask what browser they're using.

	29:34	And we started noticing that it was something specific to at the time, I forget which browser, but it wasn't ie, it wasn't Internet Explorer or Edge as it's called now. Right.

	29:46	It was something else.

	29:47	And that was great for the development team because then they found out that it was just a different way, that they had a structure, the code on how it's handled with one browser versus the other. Yeah.

	30:00	Examination of Specific Browser Issues

	30:02	So let me see if Datadog tells you what browser.

	30:11	I could have sworn that it did.

	30:15	It probably is like a Mozilla test.

	30:24	Okay. Oh, no, it does it from Edge, Chrome and Firefox.

	30:29	No failures.

	30:34	Which browser? It was.

	31:31	Oh, okay. Browser, Mozilla or Firefox?

	32:10	Yeah, it's always Firefox.

	32:17	That's interesting.

	32:19	So every time it has taken a long time to load has been because of Firefox.

	32:27	Has been through Firefox, Edge and Chrome.

	33:50	So Ricardo just joined.

	34:14	All right, let me share this stuff.

	34:40	Hey, guys. Good morning. There we go.

	34:45	Okay.

	34:47	I'm good, thanks. Yeah.

	34:49	So hopefully we can get these funds back because, I mean, I, I don't know of any information we have that was essential to justify anything along the lines of that amount of money.

	35:01	Yeah, no, no. They even said I was.

	35:03	I explained to the guys because I called them and this and.

	35:07	No, it's fine.

	35:09	It sometimes happens. Okay.

	35:11	We apologize for the, you know, inconvenience.

	35:13	What we'll do is we'll.

	35:14	We'll refund the amount minus what you should have been charged. Okay. Okay. Okay.

	35:21	So they're like they already understand that this has happened before and they're like gonna reverse the exuberant charges.

	35:28	Yeah, the pay as you go is kind of scary because okay, you know, like the recommendation was like you should go fixed thing like choose exactly the monitor you want rather than pay as you go.

	35:40	Because then I can go in there, start turning up dials and then, you know, hundreds of thousands of dollars all of a sudden. Exactly.

	35:47	Okay, I'm going to send an email to something that we just witnessed.

	35:50	Like Benji just told us that four days ago the, the entire website and yet the data dog measurement still have still seen slowdess or so let me, let me tell you what has happened.

	36:06	So between, let's say I think it was Tuesday, Wednesday, Thursday, those three days we had the WordPress site, the homepage, all the other pages, they were all loading in sub 1 second, 0.7, 0.8 seconds.

	36:24	The best we have done before was like 4 seconds, you know, on the, on the previous instance.

	36:31	Now something happened on Friday that it got a little slower and over the weekend it got back to like four or five seconds.

	36:39	So I was checking with Benji like hey, we already had it like four days working optimally.

	36:46	So I don't know what changes were done.

	36:49	I think we had to disable WP Rocket and there were some race.

	36:56	Remember that issue I mentioned over the weekend, the race.

	37:00	So that may have had something to do with it.

	37:04	But then the site was showing these errors and then when the site got back I noticed that it was like oh, four or five seconds. What is this?

	37:11	This is what we had before.

	37:14	So it's still now two seconds.

	37:16	Optimizing Website Load Speeds

	37:18	So two seconds is better.

	37:20	But we had sub one seconds.

	37:22	All our competition looking at the top schools that have an actual good website, not the ones that don't care about the website, but if you look at the ones that do a good job on the website, they're all with videos, with the content that we have, they're all like 0.8 seconds, 0.7 seconds.

	37:39	So we had pretty much beat them.

	37:41	But now we're back to being slow like two seconds instead of sub one second.

	37:49	Got it? Yeah. And so we're looking at.

	37:51	Because I didn't know, I don't know if Carlos knew, but we didn't know that the site had moved four days ago.

	37:58	But even within the last four days, the times that the performance has slow shows, like being slow, at least on Datadog is because of a JavaScript error.

	38:11	But it's specific to Firefox when you're doing your Testing like when you're seeing the, let's say the four seconds, you're also seeing four seconds or it's just students reporting that it's slow.

	38:22	No, no, I'm doing the testing myself.

	38:24	Like I'm literally testing it like okay, dozens of times a day.

	38:27	Like in the last two weeks.

	38:29	What browser are you using?

	38:31	Okay, so I'm using mostly Chrome and I'm using hello, the lcp.

	38:40	And also look, I have installed one of those widgets on Chrome the page took to load.

	38:49	So I look at full time load, like total load and then the largest content painful.

	38:56	The largest content painful is actually excellent.

	38:58	Now it's 0.5 seconds consistently.

	39:01	Whereas before it took more than a second or a second and a half or something like that.

	39:08	But the total load unfortunately went back up more than.

	39:14	Let's see right now.

	39:16	And question, can you show us the graphic you show us?

	39:23	Yeah, yeah.

	39:29	And that's interesting because Datadog definitely doesn't show that slowness.

	39:39	So this is, this is anytime that like there was a breach in the measurement, it's always attributed to this JavaScript code, but only specific to Firefox or Mozilla browsers. Okay.

	40:00	So like I go back two days for the most part, you know, it loads fast.

	40:06	But see all these little measurements here are all triggered events.

	40:11	So say March 15th?

	40:15	March 16th. Sorry, 12. Okay, midnight.

	40:17	Let's go, let's go back.

	40:45	So let's do right here, 6pm this module is just, I guess it's a JavaScript, you know, client side JavaScript code with, with Mozilla browser. Right?

	41:27	Yeah, I copied that one to the group so they can take a look at it. Okay.

	41:33	We asked Benjamin, he wasn't too sure on what that was, but we never seen.

	41:37	I haven't seen Edge come up here.

	41:39	So I guess this could be, you know, this could definitely be something completely different.

	41:46	But even you're loading the homepage right when you're seeing the slowness.

	41:51	Yeah, I mean look, the homepage is not that slow.

	41:54	It's like two seconds.

	41:55	But we already got it down to 0.7 seconds.

	41:59	And all of our competition, we found like six, seven web pages that have a video just like us, that load a video that load all kinds of sliders just like us.

	42:08	And those competitors, they do it at 0.8 seconds, 0.6 seconds.

	42:13	So obviously there's websites that are slow because they're not optimized.

	42:17	But the ones that you can tell that they've worked on optimizing it, they're able to do what we just did, and then something messed that up like in the last three days.

	42:28	So we're still working on what it was. Got it. Okay.

	42:32	So obviously this measurement is still, it's still up, but it's not every.

	42:38	Whatever the frequency was set to.

	42:40	Okay, so we're not, you know, you're gonna miss things here.

	42:44	Of course, you know, that's just.

	42:48	So our plan includes what, like five times a day?

	42:51	Three times a day or what, for what, like $50 or $100 a month or what is our plan?

	43:00	Yeah, that's because, I mean, whatever we're going to extract from it idea is that, okay, let's extract whatever value we can from this data and then once they give us the credit, maybe we either slow it down to a very low cost or eliminate it not giving us any new data.

	43:22	That's what I was going to suggest is because if we slow down the frequency to work when you have converted, the performance was horrible. Right.

	43:33	And this was like a very point in time needed, you know, measurement.

	43:38	But if it's going to be every hour, then the reality is you're going to miss when things happen. Right.

	43:40	Monitoring Strategy and Cost Management

	43:47	You know, so at that point it's like, is it even worth. Right, right.

	43:51	But what about the database connectivity or the database slowness?

	43:54	Because this one is for the website.

	43:57	But I thought with Datadog that it was going to look at like AIU databases.

	44:02	Did we monitor any databases or it was just the website.

	44:05	So Chris was trying to get a query, you know, a query or two that we can run in here and then. Yeah, we can do that.

	44:12	So that's, that's a little different.

	44:13	Like you can run that once an hour. The probe.

	44:16	The reason why the probe was so expensive is because it was being monitored from all parts of the.

	44:24	Yeah, I saw that it was like every five minutes in 11 locations.

	44:27	And then that adds up. Exactly.

	44:29	But that's still crazy because we used to have a monitoring tool that did the same thing.

	44:34	We had four or five locations around the world and a check every five minutes.

	44:41	And that that didn't cost more than like 120amonth.

	44:46	Yeah, these folks have done 80.

	44:50	When I saw that, I was like, how in the world are they charging?

	44:53	I mean, we need to get into that business right now, you know, and all this stuff is like over open source stuff that you can do. So it's like.

	45:00	Yeah, yeah.

	45:06	But I think we'll definitely look at keeping this just to monitor the database, see, at least be able to predict how much that's going to be. Right.

	45:20	So there aren't any surprises.

	45:22	Yeah, because we want to run the queries, like the database queries, to see if it suggests something, but we just have to limit that.

	45:28	Okay, look, check it once, check it three times, but, you know, cap it at a certain amount so that it doesn't just burn resources.

	45:39	But yeah, I mean, I think the datadog is a great tool. We just.

	45:43	I think we have to use it more, like deeper into the database and the queries and I mean, we have a bunch of queries already identified that we are supposedly already optimized.

	45:57	So it would be interesting to run them here and see if they agree that we optimized it or if they tell us other things.

	46:10	I'll ask in the group for Janzaiba and Benji to share the queries with Chris so that they can be uploaded to Datadog.

	46:30	Okay, well, we got the.

	46:31	We got the refund, and we.

	46:36	We don't have a refund yet. I sent. I sent the.

	46:39	The email on the 13th.

	46:40	I haven't gotten anything, you know, anything back from them.

	46:45	So, I mean, I was talking to Rick and I'm gonna try to call them today.

	46:51	Okay, fine. Yeah, I mean, if you want to call them today or based on what Rick said, I would maybe call them tomorrow or, you know, because he says he already talked and they said, oh, don't worry that we'll do it, whatever.

	47:03	Database Monitoring and Query Optimization

	47:03	So I think ideally, since he already had that chat with whoever he spoke to, maybe give them another day or two to respond, and then we can maybe call them tomorrow or Wednesday.

	47:18	But tell Rick that we need him to be on the.

	47:20	On the call with call.

	47:22	So just decide if you want to do it tomorrow or Wednesday and let Rick know, hey, join me on the call on such and such a day.

	47:28	And then, oops, you basically select which.

	47:33	Oh, come on, man.

	47:34	What is this?

	47:36	Google Meet is taking over my.

	47:40	All right, you select which database we're running on, right?

	47:45	Then you supply your query, and you get some really cool results on even, like, locks and wait time and, you know, buffers and stuff like that that normally a DBA would want to look at.

	47:58	Right. It's able to pull. So, yeah.

	48:01	All right, so I'll look out for that.

	48:06	So then, Felipe, the page that you gave me, I mean, I guess there's no point in setting that up if it's not on this environment anymore or. What do you think?

	48:15	Yeah, because I shared with him Ricardo a landing page because he was just measuring the WordPress site.

	48:23	So since the landing page is not in a different server.

	48:27	You know, I was sharing with you, but.

	48:30	Right. So the. Yeah.

	48:31	So the landing pages are where all of the marketing money goes to, and those have not been optimized yet.

	48:39	So those are still running at 3 or 4 seconds, whereas the WordPress, we had optimized it already.

	48:46	So we want to do the same thing to the landing pages that was done to the WordPress site.

	48:51	Well, the thought that we had was, let's test against a non WordPress basic HTML page. Right. So that.

	48:59	That way, when you get into these issues, you can kind of compare and say all of our landing pages are HTML.

	49:06	And unfortunately, they're mostly old.

	49:10	I mean, they're old, like two, three years old.

	49:12	And then what happens is we copy that template, we upgrade it, we change it, but I think the core code is still somewhat old.

	49:20	And so Benji has spent some time organizing the order of how things load, and he was able to decrease maybe like 50% the load time, but it was inconsistent because we would see that optimization for a few days, but then it kind of like reverts back to the normal time for some reason.

	49:42	So we were checking to see why that's happening before implementing it for all of the landing pages, because we have hundreds of landing pages, so we want to get it right for the top five or six before we launch it for all of the landing pages.

	49:59	That's interesting. Okay.

	50:04	All right.

	50:06	All right, well, thank you. Thank you, guys. That's good news.

	50:08	Felipe's gonna call on either Tuesday or Wednesday if Datadog hasn't gotten back to us.

	50:15	Felipe will call either Tuesday or Wednesday.

	50:17	If, Rick, you could be on the call with him just to, you know, just to reiterate what you were told. Yeah, And.

	50:23	And whatever happened, and hopefully, you know, by then they'll be like, okay, we're processing the.

	50:28	The refund or whatever. Or the credit. Yeah.

	50:32	All right, all right. Just reach out. All right, thank you. Thank you so much. Thank you. Okay, bye.

	50:59	Landing Page Optimization Strategy

	56:05	Concluding Remarks and Next Steps

	50:37	(End of video)


	0:00

	0:05

	0:10

	0:15

	0:20

	0:25

	0:30