How much memory should be enough for a Revo VPS?

If a Revo installation is on a VPS, and processes a lot of images, how much memory should the VPS have? Would 1GB be enough?

Would running out of memory with that server configuration mean that something is wrong?

This completely depends on the services running on that VPS and how they are configured.

If you are using nginx with PHP-FPM and running MySQL, you will only have as much RAM for the PHP workers as is not being used by nginx + mysql. Running out of memory in a single PHP child process would be based on the memory_limit you set each worker, and how low you can set this depends on the largest amount of memory your PHP requests use in a single thread. Further, if you configure FPM to allow too many children and together they attempt to use more RAM than is available on the box, the main PHP FPM process can crash entirely, taking all the children with it.

Unfortunately, it's not trivial to configure these things, but if you multiple the number of workers times the memory_limit, you can figure out a maximum memory amount FPM could attempt to use on the box. Just try to limit that to what you really have available.

For Apache configurations using mod_php, this does not apply. I haven't used this setup in a while, and I know very little about tuning it, but perhaps others can assist if that is what you are using.

This is some kind of hybrid proxy system, with nginx fronting for apache. I'm not familiar with the configuration. Zend control panel. From what I can see with a VPS ssh login, the httpd keeps getting loaded until the system runs out of memory, then it starts all over again. Nobody can get into the server to shut it down. It's an unmanaged server, so there's no help. This is all way out of my league, though, so they're on their own for this. We just moved the site onto the server a few days ago, and it's been fine until late yesterday.

That's a complex question, as there are multiple variables. Jason is correct, it does depend on how the server is configured, but also the virtualization type. An OpenVZ server can operate with less memory than a KVM server for example. Granted, it's also possible to configure a full LEMP stack with 128MB of ram. Reduced stacks can run as low as 32MB of ram, granted this requires a good bit of tweaking to accomplish.

I would recommend a 512-1024MB VPS server personally, granted if your budget allows give it a little extra room to breath. I ran our company server with 512MB of ram for over a year, increasing it to 1024 so I could implement a ram cache. Our usage is normally around 512-768MB, depending on active load at the time. The server is Debian based running on a KVM instance - OpenVZ requires fewer resources as it doesn't run the full Operating System.

If you want I can take a look at the server and give you a few tips.

The original question is moot, it's not a lack of memory, it's in some kind of loop loading the Apache httpd daemon until it runs out of memory, shuts it down, then starts all over again.

It's some kind of hybrid proxy system, where nginx is the front-end, and Apache is the back-end. It was fine for a couple of days, then suddenly started doing this... whatever it's doing. Anyway, I'm not involved in their server administration. If I do an unmanaged server, it certainly doesn't have Plesk or CPanel on it!