Server (Dedicated Server in the Cloud?)

[Erelei]

Hey guys,

Been thinking long and hard about this - and I've decided I'd like to go forward with it. As most of you know, I work at an ISP and work as a Linux Administrator. This job has perks - such as getting the server we're on right now, for free.

This has allowed our donation's coffer to fill up to almost $600, because you guys are so damn generous, and we only really need to pay for InvisionPower Board (our forum software) and if we want, additional themes and other stuff.

However, I'd like to move to a Cloud KVM. For those who don't know - a Cloud KVM is essentially a dedicated server without the physical hardware, with dedicated CPU space, full access to the kernel of the system, and a good bit of space to play with. Further, it allows a lot of flexibility in what I can do with the MUD and integrating it with the OS and systemd, which requires kernel modifications. This means better uptimes, better logging for errors, and lastly, maybe, if I can get it to work correctly, seamless reboots (no "please wait..!"). Of course, the last bit is a ways off as it'll take some time to code correctly, but it's definitely possible. I actually played with the idea on the Code port several years ago and got a working version of a seamless reboot system, but it had hiccups because of the way processes need to be replaced with sending SIGHUP signals to PIDs.

Regardless, this is going to cost some change every year. Instead of $0 for the server per year, we'd be paying $540/yr. I'm more than willing to cash in about $75/paycheck (which is every two weeks or so) - so I can definitely pay for this myself, after the initial first year payment. However, it wouldn't hurt to get some help.

Pros for this:
 

• Faster software.
• Dedicated CPUs (we share, currently. If someone on my node causes lag, the MUD lags along with it)
• Access to the kernel.
• Stability, and new features!

Cons:

• A little bit of money.
• And me asking for sponsors, if you want to put some into the pot every month to help out.

I wouldn't mind some feedback on the idea, either. Essentially, I'd like to do it regardless of anyone wanting to help out in the next year (or now, if they want to put some $ toward the paypal), but me, Anume, and the rest of the IMMs would still be eternally grateful if you had a couple dollars to spare to help me out in the years to come!

Thanks guys.

[Twinblades713]

I don't what know most of this means, but I trust you and the staff and I'm proud of the work you guys do to keep this game alive. As long as I'm around, I'm happy to contribute when and where necessary. THANK YOU!

Edited April 4, 2018 by Twinblades713

[Manual Labour]

Your hard work is appreciated. I cant commit to anything monthly but I will be making a lump sum donation soon, the least I can do for YEARS of free fun (and the occasional rage 😏)

[Celerity]

I guess it boils down to some questions:

1. Is CPU speed/space a current bottleneck for the MUD? Will running the software faster mean anything?

2. Has lag been an issue on your ISP's server? Where would be the new location of the server be?

3. How useful is kernel access to development? Seamless reboots might be nice, but a $570/year nice for that feature? Logging is always useful (you all know I love that data!), but has it been a serious bottleneck to development?

4. Is there anything else we can allocate our donations towards?

My overall opinion is, if we can get by on free reasonably well, it would take some serious necessary requirements to warrant an upgrade. If lag exists or we face hardware limitations, then it would be necessary.

With such a small playerbase, keeping our server costs free is much more sustainable. Money is useful, so is this the most useful thing we can spend our money on? Do we have to spend our money on anything right now?

If we have nothing to spend our money on, it might be wiser to save our money for when problems inevitably come up. We might HAVE to buy things in the future and it sure would be nice to have the funds there before such an emergency.

That you are willing to put your own money into the upgrade means it is important issue and I'm glad you are bring it up for discussion. I think if you want to contribute $75/paycheck to this server upgrade, you should instead contribute the $75 to the principal of a security or other instrument. Save and earn until you have to spend.

If we have about 30 players, then this server cost equates to an increase in overhead equal to ~$20/player per year. That is fairly expensive, so the gains to the consumer should also be fairly substantial to justify that increase.

Edited April 5, 2018 by Celerity

[egreir]

570/30 /12

 

Less than 1.50US a month per player.  I wouldn't call that expensive.

[English lad]

I we are considering Server options I am also happy to make an introduction to Opie @ Funcity.

For those who don't know Opie is a rather 'influential' figure in the MUD community, he maintains Funcity domain and Servers and offers both Free and Paid hosting for MUDs as part of his commitment to the Genre as a whole.

I've worked with him as staff on other Projects, and generally just find him a very decent guy, he prides himself on his track record off having hosted multiple MUDs - for free - for decades, without complaint or problem.

[Erelei]

1 hour ago, Celerity said:

1. Is CPU speed/space a current bottleneck for the MUD? Will running the software faster mean anything?

CPU is currently shared. This means that if someone else on the node causes issues (and yes, it's happened) the mud will lag. We can't determine when this lag will happen, or for how long, either. Having our own CPU that's 100% ours at any given time will prevent circumstances like these from happening. Further, it's upgraded software and does run quicker, though it may not be noticeable while MUDding because connection speed/ping > performance and speed, usually.
 

1 hour ago, Celerity said:

2. Has lag been an issue on your ISP's server? Where would be the new location of the server be?

Three or four times, now. I keep an eye on the node while at work, and we have a couple abusive assholes on it. They fix it eventually, but it usually always comes back. I'd prefer not to have to deal with it. New location of the server would be in Texas, so Central US rather than Baltimore, MD.
 

1 hour ago, Celerity said:

3. How useful is kernel access to development? Seamless reboots might be nice, but a $570/year nice for that feature? Logging is always useful (you all know I love that data!), but has it been a serious bottleneck to development?

Seamless reboots was an example. Access to the kernel allows me to integrate functions within the OS and makes things easier for me in the long run. This price wouldn't just be about access to the kernel - it would encompass all of the Pros I listed, plus more over time. Data backups and redundancy in them running versus saving them on disk (and instead on the cloud) is also a huge plus.
 

1 hour ago, Celerity said:

4. Is there anything else we can allocate our donations towards?

Maybe in time? Who knows - we usually don't know until it comes up. Fact remains, this would be a huge upgrade, and it'd help us immensely. Playerbase numbers are small, but they're also very loyal, too. I'm going to be paying toward this on my own regardless, as it'll at least help me in the future to better my knowledge in more than just C programming. In any case, I wouldn't expect the player base to contribute a damn thing towards it - though it'd be nice if they did.

Thanks for your questions.

[Erelei]

1 minute ago, English lad said:

For those who don't know Opie is a rather 'influential' figure in the MUD community, he maintains Funcity domain and Servers and offers both Free and Paid hosting for MUDs as part of his commitment to the Genre as a whole.

I've worked with him as staff on other Projects, and generally just find him a very decent guy, he prides himself on his track record off having hosted multiple MUDs - for free - for decades, without complaint or problem.

I appreciate the thought, but I don't think you read my post. I work at an ISP already and would be using my companies software and hardware, and I get serious discounts in doing so. I don't see myself allowing a 3rd party to host anything of mine, including the MUD (which is intellectually the players, but still my property as I maintain the code).

[English lad]

I expected that would be the response - but thought i'd mention it.

Take into consideration that all of the 20 + Muds he hosts there are pretty much in the same situation - and it seems to work fine - but I appreciate its not a route you need, or want to consider at this time.

[Celerity]

1 hour ago, Erelei said:

Thanks for your questions.

Thanks for spending the time to answer the questions!

[DIemar]

Almost feel like necro posting, but as someone who worked in an ISP on different positions I feel like it needs to be mentioned: as much as it is free and in ones control, people get promotions in their career and can't be a single hands on the system anymore or just have to pass the root over, some smarty pant run audits and find something unnecessary is running on the system, another smarty pants decide that hardware is underutilized and it's OK to roll several more services over an existing one killing it's free CPU cycles and so on, so on. So while OPEX is usually something to be avoided, I personally would stick to the route of a dedicated server. I, too, can't commit to a monthly payments, but will certainly be sponsoring VDS costs with irregular ones.

[Cephirus]

I know I am a bit late to this discussion (if only in that it has been months since the last post), however I do see value in the having a dedicated server, etc. and am absolutely interested in helping to fund this endeavor.

I am hopeful this is split amongst those able and interested in contributing to this endeavor, count me in as one of them. Feel free to message me the details of what my contribution would look like.