Ahosting Logo
For sites that keep seeing 508 and 503

WordPress Hosting with Published Resource Limits

"Resource limit is reached" is not a traffic cap. It is a ceiling on how many PHP requests your account may run at the same instant, and most hosts never tell you the number. Here it is printed on the plan card, with the memory, CPU, I/O and inode limits next to it, so you can size the move before you make it.

  • 30, 40 or 50 concurrent PHP workers, published
  • 2, 3 or 4 GB of memory per account
  • LiteSpeed cache keeps cached hits off the count
  • Free migration from your current host
A resource usage panel showing published entry process, memory and CPU limits with headroom, and cached requests bypassing the counter RESOURCE USAGE WP Silver account, limits published NO FAULTS Entry processes 12 of 40 Memory 1.1 GB of 3 GB CPU 85% of 300% CACHED REQUESTS NEVER ENTER THE COUNT LiteSpeed serves cached pages at the web server layer. Only logged-in, cart, search and form requests use one of the workers above.
24 Years Trusted Since 2002
99.9% Uptime Guaranteed SLA
30-Day Money-Back Guarantee
24/7 Expert Helpdesk
Ahosting, Inc. BBB Accredited Business, A+ rating

What the error is telling you

"Resource limit is reached" is a concurrency ceiling, not a traffic cap

Shared hosting runs every account inside a CloudLinux container with a fixed set of limits. The one that produces the 508 page is the entry process limit: the number of PHP requests allowed to execute at the same moment. Three things decide whether you ever hit it.

508 Resource Limit Reached

How many requests run at once

Every uncached page view, admin action, cart update, search and form post occupies one worker for as long as PHP takes to build the response. If the ceiling is 10 and the eleventh request arrives, it is refused. A slow site hits the ceiling with far fewer visitors than a fast one.

503 on LiteSpeed

Why some hosts show 503 instead

Apache with CloudLinux refuses the request instantly with a 508. LiteSpeed queues it for up to two minutes first and only returns a 503 if a worker never frees up. Same cause, different symptom. Our shared platform runs LiteSpeed, so a short spike queues instead of failing.

What never counts

Cached pages bypass the limit

A page served from the LiteSpeed cache is answered by the web server and never starts PHP, so it uses no entry process at all. On most WordPress sites that is the bulk of traffic. The published ceiling applies to the genuinely dynamic remainder, which is a much smaller number than your page views.

Want the long version? Our guide to the 508 resource limit error on WordPress walks through the concurrency math, the LiteSpeed queue and a calculator that turns your peak traffic into a worker count.

Printed, not hidden

Every limit on every shared plan

These are the CloudLinux values applied to the account, the same numbers cPanel shows you under Resource Usage. Compare them with what your current host enforces before you move.

CloudLinux resource limits per Ahosting shared plan
PlanEntry processesMemoryCPUProcessesDisk I/OIOPSInodes
WP Bronze302 GB200% (2 cores)20010 MB/s1,024250,000
WP Silver403 GB300% (3 cores)30020 MB/s2,048400,000
WP Gold504 GB400% (4 cores)40030 MB/s3,072600,000
WooStart403 GB300% (3 cores)30020 MB/s2,048400,000

Limits apply to the whole cPanel account, not per website. WooStart is the WooCommerce plan and sits at Silver-level concurrency because cart and checkout requests are never cached. A site that needs more than 50 workers on a sustained basis belongs on a VPS, and we will say so rather than sell you Gold.

Why the same site hits the ceiling less often here

Three things that change the arithmetic

The limit is only half the story. What matters is how many of your requests have to count against it, and whether anyone else on the server can spend your allowance.

LiteSpeed with LSCache on every plan

The LSCache plugin is pre-installed and active. Cached pages, including the front page, posts and category archives, are served without starting PHP. On a typical site that removes most page views from the entry process count before you tune anything.

Your limits are yours alone

CloudLinux isolates each account. A neighbour's traffic spike, runaway cron job or broken plugin consumes their workers and memory, not yours. A busy server does not lower your ceiling; only your own requests do.

You can see the counter

cPanel shows current usage and any faults for entry processes, memory, CPU and I/O under Resource Usage, with history. If a plugin or a bot starts eating workers you see it the same day, instead of discovering it through a 508 page.

Three sizes

Pick the plan by your busiest minute

Size by concurrent dynamic requests, not by monthly visitors. Workers and memory are the deciding pair and both are marked on each card. Staging, daily offsite backups, free SSL and a dedicated IP are on all three.

WP Bronze

A site that hit the limit on a 10 to 20 worker plan elsewhere.

$2.79 /month
Save 72%

US$100.44 due today for 36 months
Renews at the same price, no increase

  • Concurrent PHP workers30
  • Memory2 GB
  • CPU200%
  • Inodes250,000
  • Websites5
  • SSD storage10 GB

Pick this if: Content sites and small WooCommerce stores whose spikes are short, where caching will absorb most of the traffic.

Choose WP Bronze

30-day money-back guarantee

WP Gold

The most concurrency shared hosting here offers.

$4.79 /month
Save 65%

US$172.44 due today for 36 months
Renews at the same price, no increase

  • Concurrent PHP workers50
  • Memory4 GB
  • CPU400%
  • Inodes600,000
  • Websites25
  • SSD storage50 GB

Pick this if: Sites with long uncached requests, large catalogs or several busy sites in one account. Beyond this, a VPS is the honest next step.

Choose WP Gold

30-day money-back guarantee

Not sure which? Open a ticket with your current host's name and the plan you are on, or a screenshot of its resource usage page, and we will tell you which tier covers your peak with headroom, or whether it is a VPS question.

Moving from a host that throttles you

How the move works, and how you check it worked

Migration is free and done by our team. The point of moving is to stop seeing the error, so the process ends with a measurement, not with a DNS change.

Send us access

cPanel or FTP plus database access on the current host is enough. Keep the source site online; nothing is deleted there.

We copy and cache

Files, database and email are moved, LSCache is enabled and the site is checked on a temporary URL before you change anything.

Watch the counter

cPanel Resource Usage shows entry process, memory and CPU peaks against the published limit. Faults should read zero through your usual busy hours.

Move DNS, keep the fallback

Switch DNS when you are satisfied. The old host stays reachable until you cancel it, so there is a way back if anything was missed.

Before you order

Resource limit questions we get.

Reviewed by Adnan Canturk, Founder of Ahosting with 24 years in web hosting. Last updated:

It means your account reached one of its CloudLinux limits, almost always the entry process limit: the number of PHP requests allowed to run at the same instant. It is not a monthly bandwidth or visitor cap. A page that takes two seconds to build occupies a worker for two seconds, so slow uncached pages hit the ceiling with far fewer visitors than fast cached ones.

30 on WP Bronze, 40 on WP Silver and 50 on WP Gold, with 2, 3 and 4 GB of memory and 200, 300 and 400 percent CPU respectively. WooStart, the WooCommerce plan, matches Silver at 40 workers and 3 GB. The full table including process, I/O, IOPS and inode limits is published on this page and the same values appear in cPanel.

No. A page served from the LiteSpeed cache is answered by the web server without starting PHP, so it uses no entry process and no PHP memory. Only dynamic requests count: logged-in sessions, cart and checkout, search, form submissions, REST and admin-ajax calls and anything a plugin excludes from the cache.

LiteSpeed, which our shared platform runs, queues requests when the ceiling is reached and only returns a 503 if no worker frees up within its queue window. Apache with CloudLinux refuses immediately with a 508. The underlying cause is the same, but on LiteSpeed a short spike is absorbed rather than refused.

No. CloudLinux gives each account its own limits for CPU, memory, entry processes, processes and I/O. Another account can exhaust its own allowance without touching yours. Your ceiling is only ever consumed by requests to your own sites.

Look at the peak concurrent dynamic requests your site produces, not the monthly visitor count. Your current host may show faults in its resource usage page; the number of simultaneous entry processes at the moment of the fault is the figure to beat. If you cannot find it, open a ticket with the host and plan name and we will recommend a tier. The 508 guide linked above also has a calculator.

It buys headroom, not a cure. A plugin that holds a worker for several seconds per request will eventually fill any ceiling. The move is worth it when the site is reasonably built and simply outgrew a small plan; if the problem is one slow query or an unthrottled bot, fixing that comes first, and we will tell you when that is what we see.

When the site needs more than 50 concurrent PHP workers for sustained periods, needs Redis or Memcached, or runs long background jobs. That is a VPS workload. We say so before you order, because a customer who hits 508 on Gold a month later is not a customer we keep.

Ready when you are

Know the ceiling before you hit it

Every limit is printed, the cache keeps most of your traffic off the counter, and the migration is free. If your site is genuinely too big for shared hosting we will tell you that too.

  • Free migration by our team
  • Limits visible in cPanel
  • 30-day money-back guarantee

Published-limit WordPress hosting from

$2.79 /month

WP Bronze on a 36-month term. Renews at the same price, no increase.

Prices exclude applicable taxes. Refunds are governed by the current Terms of Service.