Nine Minecraft Plugins Most Servers Should Be Running

Most Minecraft servers run twenty plugins and need nine. The other eleven are things somebody installed once, nobody has used since, and every one of them is a future breakage after an update.

This is the short list: what a server needs, in the order it needs it, and the categories where adding one is usually a mistake.

1. A permissions manager

LuckPerms, and there is no serious alternative.

Everything else that has ranks, staff tiers or per-group settings hooks into it. Installing it first and building around it is considerably easier than retrofitting when three plugins each have their own permission model.

The specific thing it buys you is bounded staff access, a moderator who can mute and kick without being able to delete the world. Recruiting server staff covers why that boundary matters more than most owners assume.

2. Land protection

GriefPrevention, Lands or an equivalent, unless you are deliberately running anarchy.

This is the plugin that decides what kind of server you have. With it, building is safe and creativity has room. Without it, the game is about hiding.

Whichever you pick, decide the claim size and whether claims expire before launch. Retrofitting an expiry policy onto an established server is a fight; having always had one is unremarkable.

3. Essentials-style commands

EssentialsX or a modern equivalent.

Homes, warps, spawn, teleport requests, basic moderation commands, an optional economy. Almost every server ends up wanting most of this, and running it as one plugin is better than five that overlap.

The one decision worth making deliberately is whether to have teleports at all. A server with /home is a set of destinations; a server without it is a map where distance costs something. Both are legitimate and the choice shapes everything.

4. Performance tooling

Spark, and it is the most underused plugin in the ecosystem.

It profiles the server and tells you which plugin is spending the tick budget. Ten minutes with it beats a week of disabling things, and it is the step that resolves most slow servers — profiling a slow game server covers the method and why your Minecraft server lags covers what it usually finds.

Install it before you need it, not during an incident.

5. An anti-cheat

Rarely advertised, always necessary once strangers can join. The absence shows up as flying players nobody removes, and it is one of the fastest ways to lose a population.

Which one depends on your version and budget; the important part is having one and keeping it updated, since an anti-cheat that stopped updating is the usual cause of a cheating wave.

6. Logging and rollback

CoreProtect or similar.

It records who placed and broke what, and it lets you undo it. Two uses: settling disputes about what happened, and recovering from griefing without restoring a whole backup.

This is the plugin that turns "somebody destroyed my base" from a judgement call into a lookup, and it is why recovering from a griefing incident is a ten-minute job on some servers and a lost afternoon on others.

7. World management

Multiverse or an equivalent, if you want more than one world.

The specific thing worth having is a resource world that resets: a separate map people mine in, regenerated periodically. It keeps the main world from being strip-mined into something unpleasant to look at after a year, and it removes most of the pressure for a full reset. Seasons and resets covers the alternative.

8. A world border

Frequently built into your server software rather than a plugin, and it belongs on this list anyway.

An unbounded world is unbounded disk, unbounded chunk generation and unbounded backup size. Set a generous border and pre-generate inside it. This is the single cheapest performance decision available and almost nobody makes it early.

9. Chat management

Formatting, channels, a filter, and anti-spam. Which plugin matters less than having one, because unmoderated chat is where most first impressions are formed.

What not to install

A second plugin that does what one you have already does. Two economy plugins, two permission systems, two chat managers. This is the most common cause of confusing behaviour on an established server.

Anything with a "lag fix" claim. Performance comes from configuration and from removing work, not from adding a plugin. The ones that help are the ones that let you tune what already exists.

Minigames on a survival server, unless you have the population for them. A parkour course nobody uses is content in every new player's download and a plugin to maintain.

Anything from a link in a Discord message. A plugin runs with the server's permissions and can do anything the server can. Official resource sites only, game server security basics covers why this is the most likely route into a small server.

Anything you cannot say the purpose of. If it is installed and you do not know why, it is pure risk.

The maintenance rule

Every plugin is a thing that can break after a game update, and the update cycle is when unmaintained ones announce themselves. Updating without breaking things covers the discipline; the plugin-specific version is short:

Prefer maintained. Between two plugins that do the same job, recent activity beats feature count. A plugin whose last release was three years ago will break and nobody is coming.

Keep a written list. Plugin, version, source, and why it is there. Most owners cannot produce this, which turns every update into an audit.

Remove what nobody uses. An afternoon once a year, and it is the cheapest reliability improvement available.

The ones people install and regret

Shorter list than the useful one and more expensive to get wrong, because removing a plugin that stored data is harder than adding it.

Anything that adds a currency you have not designed. An economy plugin installed because economies are normal produces a currency with no sink, infinite inflation within a month, and a shop nobody uses. Server economy design covers what a currency needs before it exists, and the answer is usually that you do not need one.

Ranks with gameplay attached, sold or otherwise. A rank that grants commands is a permissions problem forever, and a rank that grants advantage is the beginning of a pay-to-win server whether you intended it or not. Cosmetic ranks cost nothing and cause nothing.

Enormous all-in-one plugins. The ones that bundle forty features into a single jar are convenient on day one and immovable on day two hundred, because half your server now depends on something you cannot replace without losing the other half.

Minigame plugins on a survival server. These come from wanting more to do and they fragment a population that was already small. A survival server with six people online does not need a parkour arena; it needs six more people.

Anything unmaintained that touches world data. A dead plugin that draws a hologram is a nuisance. A dead plugin that owns your claims, your chests or your player data is a migration you will have to perform under pressure when it finally breaks.

Installing without breaking things

The discipline that separates a stable server from one that is down every Tuesday.

One plugin at a time, with a restart and a look between each. Installing six and finding a problem means finding which of six, which takes longer than installing them separately would have.

Read the config before the first start, not after. Most plugin problems are default settings that suited somebody else's server. The defaults for an economy plugin, a claim plugin or an anti-cheat are frequently wrong for a small community.

Check the version against your server software and Minecraft version. This is the most common failure and the error is usually obvious in the log. Reading game server logs covers finding it.

Back up first, every time. Plugins that touch world data can damage it, and the window between installing something and noticing is sometimes days. Game server backups.

Test on a copy for anything that stores data. Claims, economies, inventories, homes. A copy of the world on a local server costs ten minutes and catches the failures that matter.

Keeping the stack small

The argument nobody makes and the one that holds up over years.

Every plugin is a thing that can break, a thing that must be updated, a config somebody has to understand, and a share of the tick budget. Twelve plugins is a server one person can maintain. Sixty is a job.

The performance cost is real and it is concentrated. One plugin doing work on a hot event costs more than thirty plugins doing nothing, and finding it requires a profiler rather than a guess. Profiling a slow game server covers naming the culprit in a minute.

The maintenance cost is worse than the performance cost. A Minecraft update arrives, and the number of plugins waiting on their authors is the number of weeks you are stuck on the old version. Servers with small stacks update in an afternoon; servers with large ones update in a season, or never.

Every year or so, list what is installed and ask what each one does. The answer for at least two of them will be that nobody remembers, and those two can go.

The honest shape of a good server

Nine plugins, all of them maintained, all of them with a reason somebody can state.

Not twenty, three of which are broken and nobody has noticed.