Turkey (Turkish)
Germany (German)
Worldwide (English)
PHP layerLSAPI, PHP-FPM or lsphp processes
Halkali merkez neighborhood fatih st ozgur apt no 46 , Kucukcekmece , Istanbul , 34303 , TR
When the website returns 503 Service Unavailable, the issue is often the connection between Apache and the PHP processor, PHP-FPM pool, CloudLinux LVE limit, or intense PHP processes. This expert guide diagnoses single site and entire server scenarios separately.
[proxy_fcgi:error] AH01079: failed to make connection to backend
mod_lsapi: Connect to backend failed: connect to lsphp failed: 110
Service Unavailable
The server is temporarily unable to service your request.HTTP 503 indicates the server was unable to process the request temporarily. This does not necessarily mean the domain name or DNS is faulty; the web server may be working while unable to access the PHP backend.
If only one cPanel account is affected, first examine the PHP version, handler selection, PHP-FPM pool, .htaccess rules, and CloudLinux resource history of that account.
If all PHP sites on the server are returning a 503 at the same time, check the Apache/LiteSpeed service, PHP-FPM services, global process limit, memory pressure, file descriptors, and recent EasyApache updates.
connect to lsphp failed: 110, which usually indicates a connection timeout. The lsphp process may not be responding, being created, or CloudLinux/Limits layer may be blocking the new process.
If PHP files return a 503 error when static HTML files are opened, the issue is likely with the PHP handler layer rather than the network. This distinction prevents unnecessary DNS, SSL, or firewall interventions.
Restarting the service may temporarily open the site; however, if the error persists due to intense add-ons, faulty PHP code, bot traffic, or insufficient process limits, the issue will recur. The solution is considered permanent only when the logs and source history are verified together.
Search for the phrase you see in the email, browser, or SSH log. Each card includes meaning, probable cause, and safe initial action.
Meaning: The web server temporarily failed to process the request.
Possible cause: PHP backend is closed, process limit is full, application is locked or maintenance/restart is occurring.
Meaning: Apache mod_lsapi could not connect to lsphp process in time.
Possible cause: Indicates stuck lsphp, LVE limit, intense process, memory pressure or LSAPI connection issue.
Meaning: LSAPI has repeatedly tried to reconnect to the backend and has exceeded the wait time.
Possible cause: Application is locked for a long time, PHP processes are full or remote database/API is not responding.
Meaning: The PHP-FPM pool has reached its maximum number of concurrent child processes.
Possible cause: Sudden traffic, slow PHP requests, low pm.max_children or database latency
Meaning: Apache proxy_fcgi configured PHP-FPM socket or port could not connect.
Possible cause: PHP-FPM service closed, socket path incorrect, pool not created or permission issue.
Meaning: PHP-FPM is running but cannot find the script path as a valid file.
Possible cause: DocumentRoot, proxy mapping, symlink, incorrect rewrite or file permissions.
Meaning: New process, socket or system resource cannot be temporarily detached.
Possible cause: NPROC/EP limit, process limit, file descriptor limit, or system load.
Meaning: PHP or the web server cannot allocate new memory space.
Possible cause: Real RAM exhaustion, cgroup/LVE PMEM limit or excessive process count.
Meaning: The PHP application has terminated without producing valid HTTP headers.
Possible cause: Fatal error, timeout, segfault, permission issue or premature application termination.
Meaning: There may be handler or pool incompatibility after a package update.
Possible cause: Missing package, old FPM pool, service reload or incorrect version selection.
No records matching this expression were found.
Commands are for root access. Collect only status and logs first; Do not change the permanent setting without seeing the reason.
tail -n 150 /usr/local/apache/logs/error_log
tail -n 150 /usr/local/lsws/logs/error.log 2>/dev/null503 shows the backend, LSAPI, and proxy_fcgi lines at the time.
/usr/local/cpanel/scripts/restartsrv_httpd --status
systemctl list-units --type=service 'ea-php*-php-fpm.service' --no-pagerSeparates PHP-FPM services that are installed with Apache and whether they are working.
/usr/local/cpanel/bin/rebuild_phpconf --current
/usr/local/cpanel/bin/whmapi1 php_get_vhost_versions | head -n 80Checks the system's default handler and PHP version for domain.
ps -eo user,pid,ppid,%cpu,%mem,etime,cmd --sort=-%cpu | grep -E 'lsphp|php-fpm' | head -n 40Lists CPU-using and long-running PHP processes.
free -h
journalctl -k --since '1 hour ago' | grep -Ei 'oom|out of memory|killed process'Separates normal kill operations performed by systemd due to global RAM pressure.
lveinfo --period=1h --by-fault=any --display-username --show-all
lvetopDisplays accounts with EP, PMEM, NPROC, CPU, or IO fault.
/usr/local/cpanel/scripts/check_cpanel_pkgs --fixValidates and repairs modified or missing cPanel managed packages.
find /home/USERNAME -maxdepth 3 -type f -name 'error_log' -printf '%TY-%Tm-%Td %TH:%TM %p\n' 2>/dev/null | sort -r | headEnter USERNAME to find the latest application error_log files.
First determine whether the error is account-specific or server-wide; then sequentially verify the log, handler, source, and application layers.
Test a static file, another PHP file on the same account, and a different cPanel account to determine if the issue is with the application, account, or global service level.
curl -skI https://alanadi.com/test.html
curl -skI https://alanadi.com/index.phpMatch Apache/LiteSpeed and domain error_log records by the same timestamp. The 503 message alone is not the root cause.
tail -n 200 /usr/local/apache/logs/error_logThe selected EA-PHP version of the domain, the handler used, and the PHP-FPM status must be compatible with each other.
/usr/local/cpanel/bin/rebuild_phpconf --currentIf the global issue is confirmed, restart the related PHP-FPM service and then Apache using cPanel scripts.
systemctl restart ea-php83-php-fpm
/usr/local/cpanel/scripts/restartsrv_httpdIf there is an EP/NPROC/PMEM fault, fix bot traffic, slow query, cron, plugins, and long-running PHP processes instead of just increasing the limit.
lveinfo --period=1h --by-fault=any --display-username --show-allCheck the package integrity after the update started. Then, check the HTTP code, service status, and log repeat.
/usr/local/cpanel/scripts/check_cpanel_pkgs --fix
curl -skI https://alanadi.com/First, examine the LVE history, PHP version, .htaccess, application error_log, and plugin/cron processes of that account. Global Apache restart should not be the first option.
lveinfo --user=USERNAME --period=1h --display-username --show-allApache/LiteSpeed, all EA-PHP FPM services, global RAM, process number, and last EasyApache actions are checked.
systemctl --failed --no-pager
free -h
ps -e --no-headers | wc -lNetwork and web server are basically accessible; handler, PHP-FPM socket, lsphp or PHP packages are primary suspects.
echo '<?php echo PHP_VERSION;' > /home/USERNAME/public_html/eka-php-test.phpThis situation is not a permanent solution. Long-running requests, wp-cron, attack traffic, database waits, or process limits need to be investigated.
ps -eo user,pid,%cpu,%mem,etime,cmd --sort=-etime | grep -E 'lsphp|php-fpm' | headPHP processes can fill the pool because they are waiting for the database response. MariaDB processlist, slow query log, and disk IO are checked.
mysqladmin processlist
mysqladmin status
iostat -xz 1 3HTTP 503 indicates a temporary unavailability; however, if the root cause is not resolved, it can recur indefinitely. A service restart only clears stuck processes, and does not resolve issues with intensive code or resource limits.
Indicates that mod_lsapi was unable to connect to the lsphp backend within the specified time. Investigate stuck lsphp processes, CloudLinux limits, high traffic, or memory pressure.
You can temporarily clean up attached links. However, if the error is specific to an account, a global restart may unnecessarily affect other customers and the actual cause may reappear shortly.
Only the pool is actually filling up and there is enough RAM on the server to handle each new PHP process, then it should be increased. First, slow requests and database waiting times should be detected.
DNS and SSL issues produce different error screens. A 503 response usually indicates that the request reached the web server but could not be processed at the back end.
508 usually directly indicates that the LVE resource has been exceeded. 503, on the other hand, is a more general result of backend connection or service unavailability; LVE limit may also be underneath
Your account's LVE history should be checked in public_html/error_log, wp-cron, security plugins, cache plugins, and long-running PHP processes.
Apache or LiteSpeed, domain name PHP-FPM service, MariaDB, and CloudLinux LVE layer are the basic controls. The log should indicate which layer to intervene in.
By analyzing cPanel, WHM, CloudLinux, LiteSpeed, MariaDB, Exim and security layers together, we fix the root cause of the failure instead of just removing the service.