502 bad gateway
graham88 2 yearsHi,
I keep getting a Cloudflare 502 bad gateway message.
I've noticed recently when a bingo board starts and it's a popular time for people to play, but normally a refresh sorts it. Yesterday no luck and couldn't view the website all day time UTC time.
Do others get this?
michael24 1 year
I actually open them in batches of 12 videos max when hunting for cards in a particular set and I am beginning to suspect the very obvious and that is that I am part of the problem. Cards breed traffic, traffic breeds congestion, congestion breeds hosting bills. What other ways can we think of that make the card hunting still fun and rewarding but with less traffic loads? I am not aware of how exactly the current setup and infrastructure are configured, but shooting from the hip I would think a more headless and cache supported approach might reduce traffic with a separate call to a card generator API or whatever (and yes, that will need to be smart enough to inhibit abuse). Needless to say I haven't even thought about the effort it would take to come up with a proper plan for implementation, the actual coding, validation/verification/QA and rolling out, but if I am more than willing to do what I can and if that is come up with 10 ideas of which 9 are absolute rubbish, maybe that helps.
I actually open them in batches of 12 videos max when hunting for cards in a particular set and I am beginning to suspect the very obvious and that is that I am part of the problem. Cards breed traffic, traffic breeds congestion, congestion breeds hosting bills. What other ways can we think of that make the card hunting still fun and rewarding but with less traffic loads? I am not aware of how exactly the current setup and infrastructure are configured, but shooting from the hip I would think a more headless and cache supported approach might reduce traffic with a separate call to a card generator API or whatever (and yes, that will need to be smart enough to inhibit abuse). Needless to say I haven't even thought about the effort it would take to come up with a proper plan for implementation, the actual coding, validation/verification/QA and rolling out, but if I am more than willing to do what I can and if that is come up with 10 ideas of which 9 are absolute rubbish, maybe that helps.
michael24 1 year 1 year
Following up to that; I know a whole load of really good FE developers who I could ask for advice and also the advances in AI are staggering. Bolt.new for instance is just mind blowing.
And being a business analyst in the world of digital agencies I am only too aware of the enormous gulf that usually exists between the ambitions of those that do not need to do the work themselves and those that actually have to do it. And all of it is (and should be) constrained by time and money.
Following up to that; I know a whole load of really good FE developers who I could ask for advice and also the advances in AI are staggering. Bolt.new for instance is just mind blowing.
And being a business analyst in the world of digital agencies I am only too aware of the enormous gulf that usually exists between the ambitions of those that do not need to do the work themselves and those that actually have to do it. And all of it is (and should be) constrained by time and money.
michael24 1 year
Just out of curiosity; I did some checking with Inspect in the browser to see what is cached and what not and noticed that with caching enabled, the average video detail page transfers around 380kB, but a big chunk of that is the prev.jpg. What is the purpose of that file? The video's preview image comes in at about 55kB
Just out of curiosity; I did some checking with Inspect in the browser to see what is cached and what not and noticed that with caching enabled, the average video detail page transfers around 380kB, but a big chunk of that is the prev.jpg. What is the purpose of that file? The video's preview image comes in at about 55kB
Bloopy 1 year 1 year
I've noticed the card images thing a lot too. I wondered if it was simply to do with the sheer volume of cards and/or the ways that modern browsers try to be a little too smart about caching, but I'm happy to be proven wrong if it's looked into. A second idea would be to host card images on a separate site / content delivery network. Either would do wonders for the performance of the site.
I've noticed the card images thing a lot too. I wondered if it was simply to do with the sheer volume of cards and/or the ways that modern browsers try to be a little too smart about caching, but I'm happy to be proven wrong if it's looked into. A second idea would be to host card images on a separate site / content delivery network. Either would do wonders for the performance of the site.
admin 1 year 1 year
Ultimately, we are trying to operate a website that is more popular than some medium-sized businesses on nothing more than community donations. When you see enterprises spending tens of thousands of dollars a year (for a static site!) to support a visitor count the size of ours, and we're trying to do it on (for example) the $53 we collected over the past month.
Can things be better - absolutely, we have performance stats (LCP 75th) that are 3500% above what is considered acceptable.
We could remove all the interactivity like collector cards. We could only host the current season of videos. We could make the whole site members only, or donators only. Instead;
As anyone who has been around here for a while probably notices, nearly all my time is dedicated to developing the site now (admin video uploads are almost a rarity!). Over a million lines of code was pushed to the code base over Oct-Nov dedicated to performance efficiencies. A whole new web server environment was stood up to test success with a different provider. A new upload tool (12 months in the making) is currently in testing that takes all the video processing off of the web server.
It's (unfortunately) pretty simple. With more money would come more resources, with which we would have better performance. In the current cost of living tightening across the bulk of countries that DarcyF1 members are from, it is unlikely that we're going to see a massive increase in donations, let alone the sort of numbers we'd need to provide (for example) <1s page load performance that one might see as 'normal'.
Ultimately, we are trying to operate a website that is more popular than some medium-sized businesses on nothing more than community donations. When you see enterprises spending tens of thousands of dollars a year (for a static site!) to support a visitor count the size of ours, and we're trying to do it on (for example) the $53 we collected over the past month.
Can things be better - absolutely, we have performance stats (LCP 75th) that are 3500% above what is considered acceptable.
We could remove all the interactivity like collector cards. We could only host the current season of videos. We could make the whole site members only, or donators only. Instead;
As anyone who has been around here for a while probably notices, nearly all my time is dedicated to developing the site now (admin video uploads are almost a rarity!). Over a million lines of code was pushed to the code base over Oct-Nov dedicated to performance efficiencies. A whole new web server environment was stood up to test success with a different provider. A new upload tool (12 months in the making) is currently in testing that takes all the video processing off of the web server.
It's (unfortunately) pretty simple. With more money would come more resources, with which we would have better performance. In the current cost of living tightening across the bulk of countries that DarcyF1 members are from, it is unlikely that we're going to see a massive increase in donations, let alone the sort of numbers we'd need to provide (for example) <1s page load performance that one might see as 'normal'.