[Omniscient]

It's been over a week now. How is this progressing?

It's going well. Server itself is running well. Comfortable with the security. I'm in the optimizing and customizing phase of confs. Also adding and going through my existing server customizations.

But once again I've had personal emergencies to deal with. They just don't ever seem to end. Very hard to have multiple days strung together of actual work. I'll be happy for next week.

[Dotexe]

It's going well. Server itself is running well. Comfortable with the security. I'm in the optimizing and customizing phase of confs. Also adding and going through my existing server customizations.

But once again I've had personal emergencies to deal with. They just don't ever seem to end. Very hard to have multiple days strung together of actual work. I'll be happy for next week.

Thanks for the hard work Omni! Take your time when it comes to personal things, don't feel obligated to make all that as soon as possible

[Phantom_NiNja]

It's going well. Server itself is running well. Comfortable with the security. I'm in the optimizing and customizing phase of confs. Also adding and going through my existing server customizations.

But once again I've had personal emergencies to deal with. They just don't ever seem to end. Very hard to have multiple days strung together of actual work. I'll be happy for next week.

Sounds good... take your time. I have had my share of personal emergencies... life happens.

[Komodo]

You have some controversial takes brother.

Holy hell, he lives.

[Hexicidal]

Why not just rollout to a new build completely?

Node isn't crazy. You've been rolling PHP for years. PHP is great still and with your knowledge I recommend you genuinely stay with it and just optimize your SQL and DB.

Another consideration is build off a different forum platform with JS as the focus. A React upgrade would take you less than a year if you took it seriously. I know the fear of knowing JS is a mindkiller, but the fundamentals are the systematically the same. Then React flexes the power of state management on crack. The API can be rebuilt and utilized to service the DB correctly.

You can even split this further and just run a split test server and half migrate. Run your worries and then see if it works. I'm not going to help you build your site, but if you want a path deeper I can help pick your brain. Whatever you want.

[Cooligan]

Why not just rollout to a new build completely?

Node isn't crazy. You've been rolling PHP for years. PHP is great still and with your knowledge I recommend you genuinely stay with it and just optimize your SQL and DB.

Another consideration is build off a different forum platform with JS as the focus. A React upgrade would take you less than a year if you took it seriously. I know the fear of knowing JS is a mindkiller, but the fundamentals are the systematically the same. Then React flexes the power of state management on crack. The API can be rebuilt and utilized to service the DB correctly.

You can even split this further and just run a split test server and half migrate. Run your worries and then see if it works. I'm not going to help you build your site, but if you want a path deeper I can help pick your brain. Whatever you want.

I heavily agree with the React upgrade part. React has been known as the "app-like" language in terms of overall site loading speed due to its memory capability (sophisticated state management). Frontend wise, this would be a huge upgrade. HF already loads fairly fast for a forum, especially when compared to other popular MyBB forums. Apart from the initial first slow-load after upgrading due to React having to download new code, once it's installed, the forums interactivity and overall load times will function at app-like speeds.

As for the backend, optimizing SQL would also be great here. Instead of running Full Table Scans, we would have proper indexing. Also, regarding Omniscient and how he doesn't like to add extra load to the site, Query Efficiency would fix this. Instead of loading every single bit of a users information when only trying to find specific information, the optimized SQL would turn it into a more surgical process rather than a poke-and-prod style approach.

The only real downfall I predict could potentially be reduced SEO rankings. PHP is great for this, allowing Google to read site text instantly. If React isn't setup perfectly, every page will first be loaded blank, then use JavaScript once the pages content is actually delivered to load it. Because of this, Google might end up seeing empty forums instead of immediate cached content, thus killing some of HFs Google SEO in the process. This would not be good obviously and the only thing I'd be "worried" about, especially since Omni has put in very consistent work into growing HFs SEO rankings over the years.

Basically, yes you could stick to what you know and keep things how they are for the most part, the "house" that is HF has always worked well for us end-users. Instead of moving into a new house with similar ammentities though, with Hydra 's suggestion, HFs "new house" would essentially act like and be upgraded to a smart-home with optimized SQL and the implementation of React for site speeds. Loading times are everything to users nowadays since most sites tend to optimize this the most, especially social media sites that I suppose HF could fall into the category of.

I understand how this overhaul would turn a relatively short process into a fairly long one, especially due to the learning curve. I mean, as you stated in the OP, the current state of the backend is a liability. Trying to put a React frontend on an outdated and potentially partially broken backend is like trying to put a Ferrari chassis on an old rust-bucket engine, so again I understand any hesitancy. In the long run though, it would act as future-proofing for HF in terms of optimized load balance and efficiency.

[CL1B]

Cooligan Hexicidal Hell no, this is and will always be the largest bb board. The theme, the plugins everything ties into HF's branding.

That's something so valuable, you wouldn't know if you didn't own shops.

Yes React is wonderful and todays standard yeah yeah yeah, but when you already have a well written site why would you port it? The performance can't be much greater, really not. In fact, I could argue it would take a shit ton of time to get a react port of HF board to be on par with current performance. Also, from scratch is a terrible idea; We have trouble getting you guys to write threads, be active etc and you want a whole new board from the ground up?

Omni clearly stated he's using the little time he has on migrating servers, the fuck you suggesting a whole new board for?

Look, if you guys want a more modern round and flat feel to the UX suggest a new theme.

Also please do not mention the DB as if the man hasn't been optimizing that for years, you have to know HF has done more migrations in the past, more optimizations and so on. Everything is buttery smooth except for the backup connection pool.

The server move is for headroom, so we can implement more features and have a more headroom whilst doing so. There's 1 thing that could seriously benefit from a rewrite or new version and that would be the v2 api, I believe there's already some momentum for this with user like xadamxk offering their expertise for it.

[xadamxk]

Why not just rollout to a new build completely?

Node isn't crazy. You've been rolling PHP for years. PHP is great still and with your knowledge I recommend you genuinely stay with it and just optimize your SQL and DB.

Another consideration is build off a different forum platform with JS as the focus. A React upgrade would take you less than a year if you took it seriously. I know the fear of knowing JS is a mindkiller, but the fundamentals are the systematically the same. Then React flexes the power of state management on crack. The API can be rebuilt and utilized to service the DB correctly.

You can even split this further and just run a split test server and half migrate. Run your worries and then see if it works. I'm not going to help you build your site, but if you want a path deeper I can help pick your brain. Whatever you want.

While I love React, and do think it is the path forward for web dev, I don't think moving to it adds much value to the site given the amount of work it would take.
He could focus on other plugins and features that bring activity or users with a fraction of the investment.

[Hackers]

While I love React, and do think it is the path forward for web dev, I don't think moving to it adds much value to the site given the amount of work it would take.
He could focus on other plugins and features that bring activity or users with a fraction of the investment.

yeah full conversion would take a while if you want to keep all of the customizations.
Tough position to be in

[Nachfolger]

This thread has gone off the rails. Why don't you guys quit yapping about completely unrelated BS and let the guy finish the migration in peace.

[Transporter007]

The situation is this. HF needs to upgrade our entire backed. Upgraded PHP, MySQL, Nginx, Apache, NodeJS, and the OS. It's vital because without the upgrade of the server we can't really do modern development and improvements to HF. Our backend is out of date, so much so it's now becoming a liability. I can't do the integrations and usage of new libraries without this.

I have a new server with excellent specs already provisioned. I now am faced with the task of building up the new server. The last time I did this seems like a decade ago or more. I do have anxiety about the process but I have no choice, this must get done. I'm choosing this week to really push through because my wife is on an 8 day vacation and this is my chance to work without family interruption.

This is my workflow plan.

1. Install OS.
2. Install Control Panel and services.
3. Customize server for security and special needs of HF.
4. Move HF to new server (but not change DNS records).
5. Use my Hosts file to point my computer to new server as HF.
6. Begin working to update the site.
7. Once HF is functional on new server, change DNS for everyone.
8. Continue fixing and upgrading anything broken (which is going to be a lot).


Very likely we're going to lose Convo. I just expect this. It's not the highest priority given the overall forums is more important. I won't give up on Convo but it's also just one feature and I'm fairly sure we can live without it for a week or more. Fixing on-site errors (PHP, MYSQL) will take priority.

I plan on changing how our web socket works which is notify and convo. Fairly sure I'll be able to fix notifications quicker than Convo. It's a simple system. Or I'll just rebuild it from scratch. They are actually integrated and going forward I may attempt to completely separate them. Both run under NodeJS as app.js but would be better imho as notify.js and convo.js instead. Currently the socket is on a different port. Nginx does have a proxy forwarding for WS but as I stated, our version of Nginx is outdated and doesn't support that. This causes some strange issues like Cloudflare blocking the port because of not passing Captcha, which doesn't ever display on web socket. Integration of the WS to port 443 (HTTPS) will fix that ongoing issue.

I will be using AI through much of this. It's a very useful tool for coming up with configuration files like my.cnf (MySQL Configuration) and firewall (UFW/IPtables) rules. And then I have to consider if I'll install an Agent Assistant to go through all of HF's codebase to update to PHP8. That's going to be very experimental for me.

The good news, I've done this a few times, the bad news, not with NodeJS which is where I anticipate most of my struggling will come from. Some members have offered to help but I am not relying on that. Forum member help is always unreliable and in the end, it's best that I have full control and understanding of how HF works. If I can build it and fix it, then it's always going to be part of HF.

I'm both confident I can do this and apprehensive about the unknown. But this must be done and we're long past due. I'll keep you all posted as I progress. It will be point 5 where I'll feel most energized. That's when the real work begins.

Thank you.

I still think HF Database Size would be ?

[Omniscient]

I still think HF Database Size would be ?

It's about 38GB currently.

[Jamika]

This thread has gone off the rails. Why don't you guys quit yapping about completely unrelated BS and let the guy finish the migration in peace.

That’s exactly the point of the topic—otherwise, he wouldn’t have posted it in the first place.

[Transporter007]

It's about 38GB currently.

Its alot man 😲

[Nachfolger]

That’s exactly the point of the topic—otherwise, he wouldn’t have posted it in the first place.

Yeah for sure. He definitely posted this thread so you people can suggest he do a complete migration to a React.js frontend. You got me there. 👍