Panellicense

Fix 508 Resource Limit Is Reached on CloudLinux cPanel

A 508 error means one LVE limit got hit — usually entry processes. The decision tree for finding which limit, which user, and whether to raise it or fix the cause.

5 min readUpdated 2026-06-05error · cloudlinux · lve · 508
schema: TechArticleschema: FAQPageschema: BreadcrumbList

A visitor sees "508 Resource Limit Is Reached" in the browser; the site owner opens a ticket. The message means CloudLinux refused to place an Apache/PHP request into the user's LVE because a limit was already maxed out. It is almost always the entry processes (EP) limit — but not always, and raising the wrong limit just moves the problem.

This is the order to diagnose it. Work top to bottom; stop when the data points at a cause. The audience is the host, not the end user — you have root.

1. Confirm it's LVE, not something else

508 is a CloudLinux-specific code emitted by mod_hostinglimits. A real 508 always leaves a fault record. If lveinfo shows zero faults for the period the user complained about, you're chasing the wrong error — check for a 503 from PHP-FPM or a WordPress fatal instead.

# Faults across all users in the last day, sorted by entry-process faults
lveinfo --period=1d --by-fault=mep --display-username

If the user's name shows up with a non-zero count, that's your 508. Note the number — it's how many requests were served a 508 in that window.

2. Find which limit is actually faulting

Don't assume EP. Pull the per-fault breakdown so you raise the right knob:

lveinfo -d --period=2h \
  --show-columns="ID,CPUf,EPf,PMemF,NprocF,IOf,IOPSf" \
  --display-username

Read the column with the non-zero count:

  • EPf — entry processes. Default is 20 concurrent requests. By far the most common 508.
  • NprocF — total processes/threads (default 100). Points at a fork bomb, a stuck cron, or exec() loops.
  • CPUf — CPU speed (default 100% = one core). The site is genuinely compute-bound.
  • PMemF — physical memory (default 1 GB RSS). A heavy plugin or a low memory_limit mismatch.
  • IOf / IOPSf — disk throughput / operations. Usually an un-cached site hammering MySQL on disk.

For the exact moment a fault fired, read the snapshot — it records the offending PIDs and their resource usage:

lve-read-snapshot -u <username>

3. Decide: raise the limit, or fix the cause

A 508 is a symptom. Before you raise anything, ask whether the load is legitimate.

Raise the limit when the user is genuinely busy and on a plan that should allow it — a real traffic spike, a launch, a logged-in WooCommerce checkout crowd. Edit in WHM under Server Configuration → CloudLinux LVE Manager, find the user, click Edit, and bump the faulting limit. Or from the shell:

# Raise entry processes for one user to 40
lvectl set <username> --ep=40 --save

# Confirm
lvectl list | grep <username>

Fix the cause when the load is waste — and most chronic 508s are waste:

  • No page cache. An un-cached WordPress site hits PHP+MySQL on every request and burns EP under any bot traffic. Turning on LSCache on cPanel often drops EP faults to zero with no limit change.
  • Bot/crawler floods. Check the domain's access log for one aggressive user-agent or xmlrpc.php/wp-login.php hammering. Block it at the firewall, not by raising EP.
  • suPHP or DSO instead of lsapi. Slow handlers hold a worker open for the whole request, so concurrency piles up. Switching to mod_lsapi keeps requests short — see mod_lsapi tuning.
  • A slow upstream. A third-party API call blocking PHP for 8 seconds means 20 such hits exhaust EP instantly. That's an application bug, not a limit.

4. Set the package default sanely

If you're handing out 508s to brand-new accounts, the package default is too tight. Set realistic per-tier defaults once so you stop firefighting per-user — the full method is in CloudLinux LVE tuning, and the workflow for spotting repeat offenders is in reading LVE Manager statistics.

One cleanup while you're in there: set VMEM to 0 (disabled). Virtual memory limits are deprecated and produce confusing OOM-style faults that masquerade as other problems.

lvectl set-default --vmem=0 --save
What causes a 508 Resource Limit Is Reached error?+
One of the user's LVE limits is maxed when a new request arrives. Most often it's entry processes (EP, default 20) — too many concurrent PHP requests — so mod_hostinglimits returns 508 instead of placing the request.
How do I find which CloudLinux user is hitting 508?+
Run `lveinfo --period=1d --by-fault=mep --display-username`. Any user with a non-zero count was served 508s in that window. Use `lveinfo -d --show-columns=...` to see which specific limit faulted.
Should I just raise the entry process limit to fix 508?+
Only if the traffic is legitimate. Most chronic 508s come from missing page cache, bot floods, or a slow PHP handler. Fixing the cause is more durable than raising EP, which can let one site destabilise the whole node.
Is a 508 error the same as a 503 error?+
No. 508 is CloudLinux's LVE-limit signal from mod_hostinglimits. A 503 usually comes from PHP-FPM, Apache, or a proxy and is unrelated to LVE. If lveinfo shows no faults, you're looking at a 503-class problem instead.

Next steps

Running CloudLinux unlicensed or on the wrong tier? Sort it from your CloudLinux license page, or ask us on contact.

Related articles

All troubleshooting →
troubleshoot
Rspamd on cPanel: what's actually supported in 2026
cPanel & WHM has no native Rspamd integration — here's what the built-in stack actually is, why community Rspamd-on-Exim setups are risky, and when switching panels is the real answer.
5 min read
troubleshoot
Fix JetBackup license errors: invalid, not found, and IP mismatch
Diagnose JetBackup's license validation errors — no license found for this IP, license invalid for current server, and reissue failures — and fix each from WHM or the CLI.
5 min read
troubleshoot
Fix InterWorx license errors: invalid key and sync failures
Diagnose the Invalid InterWorx-CP License error — DNS and port 2443 checks, license.pex --sync, IP rebinding, and key swaps — before the 24-hour grace period ends.
4 min read
Switch in an afternoon

Switch from your current reseller — free.

We migrate active cPanel, Plesk, LiteSpeed and CloudLinux licenses from any reseller. We prorate the first month so you never pay twice, and your customers see zero downtime during the swap.