# "Resource limit is reached" (508 error): causes and fixes

URL: https://shrotihost.in/blog/resource-limit-is-reached-508-error
Published: 2026-11-20. By the Shroti Host team, founded by Priyanshu Shroti (https://shrotihost.in/about/priyanshu-shroti).

> A 508 "Resource limit is reached" error means your hosting account had more PHP requests running at one moment than its CloudLinux entry process limit allows, so the server refused the extra ones. It is a concurrency limit, not a visitor limit. The usual fix is a page cache and blocking bot traffic, not a bigger plan.

If your site showed "508 Resource limit is reached", or your host emailed you about resource faults, this guide is for you. It assumes cPanel hosting on CloudLinux, which a lot of shared hosting runs on, ours included, and it leans toward WordPress because that is where the error usually appears.

You will get what each of the six limits does when you reach it (from the CloudLinux documentation), the limits we publish for every plan, a way to find out which limit you hit and what was running at the time, a fix order, and a plain rule for when an upgrade is the right answer.

## What does "Resource limit is reached" mean?

"Resource limit is reached" is the page a CloudLinux server returns, with HTTP status 508, when your account already has as many requests executing as its entry process limit allows. The server could not start one more, so it answered with the error instead of your page.

CloudLinux puts every hosting account in its own container, called an LVE (Lightweight Virtualized Environment), with a set allowance of CPU, memory, disk throughput and processes. The [CloudLinux limits documentation](https://docs.cloudlinux.com/cloudlinuxos/limits/) states the purpose in one line: "The goal of LVE is to make sure that no single web site can bring down your web server." The limit that stopped your page is the same mechanism that stops another customer's traffic spike from slowing your site.

Think of the entry process limit as a row of slots. A request that needs PHP takes a slot when it starts and gives it back when it finishes. A page served from cache, an image or a CSS file does not start PHP, so it does not need a slot. The 508 appears only in the moments when every slot is taken and another PHP request arrives.

![Diagram of a hosting account at its entry process limit: cached pages and static files are served without using a slot, PHP requests fill all 20 slots, and the next PHP request receives a 508 Resource limit is reached error.](https://shrotihost.in/blog/resource-limit-is-reached-508-error-light.webp)

## The six CloudLinux limits and what happens at each one

Only one of the six limits produces a 508. According to CloudLinux, CPU and I/O limits make a site slower, memory and process limits produce 500 or 503 errors, and the entry process limit produces the 508. That matters for diagnosis: a slow site and a 508 can be two different limits.

The table gives each limit with the short name you will see in cPanel and the behavior CloudLinux documents. Some hosts also cap the number of files (inodes). That is a separate limit and does not cause a 508.

Table: CloudLinux LVE limits and what happens when an account reaches each one

| Limit | What it caps | At the limit | What you see |
| --- | --- | --- | --- |
| CPU (SPEED) | Processor time, as a share of a core or a number of cores | The account's processes are slowed to fit the allowance | Slow pages, no error |
| Physical memory (PMEM) | RAM actually used by your processes | CloudLinux kills some processes in the account | 500 or 503 errors, half-finished imports or backups |
| I/O | Disk throughput in MB per second | Processes are throttled (put to sleep), not stopped | Slow pages, slow backups |
| IOPS | Disk read and write operations per second | Reads and writes pause until the current second ends | Short stalls on database-heavy pages |
| Entry processes (EP) | Requests entering the account at once: PHP requests, plus SSH sessions and cron jobs | The web server cannot place the request and returns 508 | "Resource limit is reached" |
| Number of processes (NPROC) | All processes and threads inside the account | No new process can start until another one ends | 500 or 503 errors |

## Why does a site with few visitors still get a 508?

A low-traffic site gets a 508 because the limit counts requests in flight at one instant, not visitors per day. Two things fill the slots: more PHP requests arriving, or each request taking longer to finish. The second one is the trap.

The arithmetic is simple enough to do yourself. With 20 entry processes, if an uncached page takes 2 seconds to build, the account clears about 10 PHP requests per second before the slots are full. If the same page takes 0.2 seconds, it clears about 100. One slow database query or one plugin calling a slow outside API can cut your capacity by ten times without a single extra visitor.

Much of the traffic that fills slots is not people at all. Crawlers, login brute-force scripts and uptime checkers each hold a slot for as long as their request runs. CloudLinux also notes that entry process counting can differ by web server, LiteSpeed included, so treat the number as a ceiling on concurrent PHP work, not an exact count of connections.

## The limits we publish for each Shroti Host plan

These are the limits we publish for each plan on the [shared hosting](https://shrotihost.in/hosting) and [Unlimited Hosting](https://shrotihost.in/unlimited-hosting) pages, taken from the same plan data the site uses. WordPress plans publish the same figures as the shared tier in the same row, so moving from Launch to WP Starter changes the WordPress tooling and none of these numbers ([shared hosting vs WordPress hosting](https://shrotihost.in/blog/shared-hosting-vs-wordpress-hosting) has the comparison).

Two notes on reading it. These are ceilings on what an account may use, not capacity set aside for you. And when you are diagnosing a real problem, the figure that counts is the limit column on the Resource Usage page in your own cPanel (next section).

For what "unlimited" does and does not cover on the second half of the table, see [is unlimited hosting really unlimited](https://shrotihost.in/blog/is-unlimited-hosting-really-unlimited).

Table: Published resource limits for Shroti Host shared, WordPress and Unlimited plans

| Plan | CPU and RAM | I/O and IOPS | Entry processes | Processes |
| --- | --- | --- | --- | --- |
| Launch, WP Starter | 1 core, 1.5 GB | 5 MB/s, 1,024 | 20 | 40 |
| Grow, WP Business | 1.5 cores, 2 GB | 10 MB/s, 1,536 | 30 | 60 |
| Turbo, WP Performance | 2 cores, 3 GB | 15 MB/s, 2,048 | 40 | 100 |
| Infinity, WP Ultimate | 3 cores, 4 GB | 20 MB/s, 3,072 | 60 | 150 |
| Unlimited Solo | 3 cores, 4 GB | 25 MB/s, 4,096 | 75 | 150 |
| Unlimited Trio | 4 cores, 5 GB | 30 MB/s, 5,120 | 100 | 200 |
| Unlimited Pro | 5 cores, 6 GB | 35 MB/s, 6,144 | 125 | 250 |
| Unlimited Infinity | 6 cores, 8 GB | 40 MB/s, 8,000 | 150 | 300 |

## How do you find out which limit you hit?

The Resource Usage page in cPanel tells you which limit you hit and when. It is CloudLinux's end-user plugin, found in the Metrics section (some themes label it "CPU and Concurrent Connection Usage"). The steps below follow the [CloudLinux Resource Usage documentation](https://docs.cloudlinux.com/cloudlinuxos/lve_manager/#resource-usage-client-plugin). If your cPanel does not show the page, ask your host for the account's LVE statistics for the time of the error.

1. Open the Dashboard tab. It says whether the site was limited in the last 24 hours and names the resource. Click Details.
2. On Current Usage, read the table: one row per resource with Usage, Limit and Fault. A fault is one moment when the account asked for more than the limit. Faults on entry processes are your 508s. Faults on physical memory or processes are 500 and 503 errors. CPU or I/O sitting at the limit with no faults explains slowness.
3. Set the timeframe to the day of the problem and the unit to hour or minute. Note the exact time the faults start and whether they repeat on a schedule. Faults at the same minute every hour point to a cron job or a scheduled backup. Faults at random times point to traffic.
4. Open the Snapshot tab and pick a snapshot near that time. CloudLinux takes these after faults and lists the running processes, the database queries and the HTTP requests. The same URL appearing many times is your answer.
5. Confirm in the access log (Metrics, then Raw Access or Visitors). Look at the fault minute for repeated hits on wp-login.php, xmlrpc.php, wp-cron.php or admin-ajax.php, and for one IP address or user agent making most of the requests.

## What usually causes a 508 on WordPress?

On WordPress the cause is almost always uncached PHP, either because nothing is cached or because the traffic is the kind a cache cannot serve. These are the five we would check, in the order they tend to turn up in a snapshot:

- No page cache. Every visit builds the page in PHP and holds a slot while it does. This is the first thing to check and the cheapest to fix.
- Bots and crawlers. Aggressive crawlers request many pages at once, often search or filter URLs with query strings that no cache holds.
- Login and XML-RPC floods. Scripts posting to wp-login.php and xmlrpc.php never get a cached response, so every attempt costs a slot.
- WP-Cron. WordPress checks for scheduled tasks on page loads, and a pile of overdue tasks (backups, feed imports, email queues) can run during your busiest minute.
- Heavy plugins. Broken-link checkers, statistics plugins, related-post queries and anything that polls admin-ajax.php keep PHP busy even when no visitor is on the site.

## The fix order: what to do first

Fix in this order, because each step is cheaper than the one after it and the first two deal with the most common causes. Check Resource Usage again after each step so you know which one worked.

1. Turn on a page cache. On a LiteSpeed server, install the LiteSpeed Cache plugin and enable page caching; LiteSpeed Cache is included on every Shroti Host shared, WordPress and unlimited plan. Then open the site in a private window and confirm pages come from cache.
2. Cut the traffic that is not customers. Disable XML-RPC if you do not use it, limit login attempts, and block or slow the crawler you found in the access log. You can add Cloudflare's free CDN to your domain and apply those rules there, before requests reach the server.
3. Move WP-Cron to a real cron job. Add define('DISABLE_WP_CRON', true); to wp-config.php and schedule a request to wp-cron.php from cPanel's Cron Jobs, as the [WordPress developer handbook](https://developer.wordpress.org/plugins/cron/hooking-wp-cron-into-the-system-task-scheduler/) describes. Every 15 minutes is enough for most sites. Cron jobs count as entry processes too, so do not schedule them every minute.
4. Remove or replace the plugin the snapshot named. Deactivate it, watch the faults for a day, and reschedule backups and imports to your quietest hours.
5. Make the remaining PHP faster. Use a current PHP 8 version and, where your plan includes it, a Redis object cache (on our plans: Turbo and Infinity, WP Business and above, and every Unlimited tier). Our guide to [speeding up WordPress](https://shrotihost.in/blog/speed-up-wordpress-website) covers the rest.

## When should you upgrade, and to what?

Upgrade when the page cache is on, the junk traffic is blocked, and Resource Usage still shows faults on ordinary days from requests that cannot be cached: checkouts, carts, logged-in members, searches, an API. That is real demand, and tuning will not remove it. If the faults come from one bad day, one bot or one plugin, a bigger plan only makes the same problem cost more.

Pick the tier by the resource that faulted. For entry process faults, compare the entry process column in the table above and move to the next figure up: Launch (20) to Grow (30) is ₹49 to ₹79 a month excl. service fee. For memory faults, look at the RAM figure instead. Past the top of the [shared hosting plans](https://shrotihost.in/hosting), the larger published allocations are on the Unlimited line, which is billed monthly only: Unlimited Solo publishes 75 entry processes for one website at ₹199 a month, and Unlimited Infinity 150 at ₹499.

If a tuned site would still fault at 150 entry processes, or it needs long-running workers, queues or custom server software, shared hosting of any brand is the wrong product and you need a server of your own. Our [VPS plans](https://shrotihost.in/vps) give you root access and the vCPU and RAM listed for the plan, but you administer the server yourself, VPS plans are not refundable, and we have no India location. Read the [signals that it is time for a VPS](https://shrotihost.in/blog/when-to-upgrade-to-vps) before you move.

Table: Fix or upgrade: what each Resource Usage pattern means

| What Resource Usage shows | Likely cause | Do this |
| --- | --- | --- |
| Entry process faults on one day, snapshot full of one URL or one IP | Bot, login flood or a single plugin | Block it or remove it. Do not upgrade |
| Faults at the same minute every hour or night | Cron job, backup or import | Reschedule and move WP-Cron to a real cron job |
| Entry process faults in busy hours, no page cache yet | Uncached pages | Turn on the page cache, then measure again |
| Faults on normal days after caching, from checkout or logged-in requests | Real uncacheable demand | Move up to the next entry process figure |
| Physical memory faults during imports or image processing | One process needs more RAM than the plan publishes | Run the job in smaller batches, or move to a tier with more RAM |
| Still faulting on the largest shared or unlimited tier | The site has outgrown shared hosting | Move to a VPS |

## Frequently asked questions

### Is a 508 error the same as a 503 or 500 error?

No. On CloudLinux hosting a 508 means the account reached its entry process limit, so the server could not start another request. CloudLinux documents 500 and 503 errors as the usual result of reaching the physical memory or number of processes limit instead. A 500 can also be an ordinary PHP or .htaccess error with no limit involved, so check the Resource Usage faults before assuming.

### Can bots cause "Resource limit is reached"?

Yes, and they often do. A crawler or a login brute-force script sends many requests at once to pages that cannot be served from cache, and each one holds an entry process while it runs. The access log for the minute of the fault shows it: one IP address or user agent making most of the requests. Blocking or rate limiting that client usually ends the errors.

### Does the number of WordPress plugins trigger this error?

Not the count itself. Thirty light plugins can be fine while one heavy plugin fills the slots. What matters is how much PHP work each request does and whether a plugin runs background jobs or polls admin-ajax.php. Use the Snapshot tab in Resource Usage to see which requests were running at the fault, and judge plugins by that evidence.

### Does another website on the same server cause my 508 error?

No. The 508 is returned when your own account reaches its own entry process limit. CloudLinux gives each account a separate container with separate limits so that one site's load is contained. A busy neighbor can be a reason a shared server feels slower in general, but the fault counters on your Resource Usage page only count your account's usage.

### What is CloudLinux in cPanel hosting?

CloudLinux is an operating system widely used for cPanel shared hosting. It places each hosting account in a container called an LVE with its own limits for CPU, physical memory, disk I/O, IOPS, entry processes and number of processes. In cPanel you meet it as the Resource Usage page, which shows your usage, your limits and any faults.

### How long does a 508 error last?

Only as long as every entry process slot stays busy. Each slot is released when its request finishes, so the error clears by itself once the queue of PHP requests drains, often within seconds. Nothing needs to be reset. If the 508 keeps coming back, the cause is still running: find it in the Resource Usage snapshots and the access log for the minutes when faults were recorded.

Next step: [See the published limits on every plan](https://shrotihost.in/hosting). Shroti Host shared hosting starts at ₹49/mo excl. service fee, and each plan page lists the CPU, RAM, I/O, IOPS and process limits we publish for it. Fix the site first; upgrade only if the faults stay.
