Hey all, So I was seeing a few other people reporting issues after their control panel auto-updated, maybe it’s related to that, but… I have two separate instances of Hestia, and on each instance there are 1-2 sites that are constantly timing out. Some stuff I’ve ruled out: Both are behind Cloudflare, so it’s not bot traffic Both have unlimited bandwidth allocated Other sites on the server are unaffected Neither have received any recent code changes (so it’s not on the website end) Tried increasing php-fpm settings for the site with a custom template (didn’t change anything) Checked for any long-running scripts, like a broken php artisan schedule:run or w/e (didn’t see anything) The servers are both Hetzner VPSs with 4 vCPUs, 16 GB RAM, 2 TB traffic so I really don’t think it’s resources The error.log is amazingly vague (all my sites use NGINX): [error] 1287#1287: *20612 connect() to unix:/run/php/php8.3-fpm-DOMAINNAMEHERE.net.sock failed (11: Resource temporarily unavailable) I have no idea where to go from here. I don’t want to just increase the timeout limit because that’s not really a solution, it’s more of a bandaid. (and I think cloudflare has their own timeout limit anyways) Does anyone have any suggestions? Thanks!

Hi, [image] perabyte: Both are behind Cloudflare, so it’s not bot traffic Are you sure? Take a look to this topic just in case: [image] High Cpu Usage - Spammed by meta? I seem to be getting spammed by these 57.141.16.24 - - [20/Apr/2026:04:00:50 +0800] "GET /cart/?remove_item=bfd2308e9e75263970f8079115edebbd&_wpnonce=87af8e445c&add-to-cart=4571 HTTP/2.0" 200 1452108 "-" "meta-externalagent/1.1 (+https://developers.facebook.com/docs/sharing/webmasters/crawler)" 57.141.16.45 - - [20/Apr/2026:04:00:50 +0800] "GET /cart/?remove_item=a927541d74113c53810fb0b383a649ac&_wpnonce=c5368d0313&add-to-cart=4571 HTTP/2.0" 200 1452108 "-" "meta-externalagent/1.1 (+https://dev… [image] perabyte: Other sites on the server are unaffected It looks like those sites are exhausting the pm workers. [image] perabyte: Tried increasing php-fpm settings for the site with a custom template (didn’t change anything) What did you try and what is the current php-fpm conf for those sites? [image] perabyte: The error.log is amazingly vague (all my sites use NGINX): [error] 1287#1287: *20612 connect() to unix:/run/php/php8.3-fpm-DOMAINNAMEHERE.net.sock failed (11: Resource temporarily unavailable) Did you see any max_children message in the logs? journalctl -u php8.3-fpm | grep max_children

[image] sahsanu: Are you sure? Take a look to this topic just in case: Yeah, I scrolled through the nginx logs and I didn’t see any unexpected traffic or user agents. I have a very thorough robots.txt on top of cloudflare “under attack” mode on – all the requests look pretty legitimate. [image] sahsanu: What did you try and what is the current php-fpm conf for those sites? The php-fpm conf was previously: (this should be hestia default because I haven’t touched it before now) pm = ondemand pm.max_children = 8 pm.max_requests = 4000 pm.process_idle_timeout = 10s Some pm settings I’ve successfully used outside of Hestia for higher traffic sites was this, so I tried changing it to that: pm = ondemand pm.max_children = 80 pm.max_requests = 200 pm.process_idle_timeout = 10s I’m definitely not a ondemand config wizard so if you have a better idea of settings to try, let me know! [image] sahsanu: Did you see any max_children message in the logs? No, I ran that command and it came back empty. I looked at the journalctl for php8.3-fpm in general and I didn’t see any abnormal messages.

[image] perabyte: pm = ondemand pm.max_children = 80 pm.max_requests = 200 pm.process_idle_timeout = 10s I’m definitely not a ondemand config wizard so if you have a better idea of settings to try, let me know! The right conf will depend on the RAM used by each of the php workers and the traffic on those sites. Show the output of these commands: ps --no-headers -e -o "pid,rss,cmd" | grep -E "[p]hp-fpm: pool" | awk '{printf "PID %s: %.1f MB RSS\n", $1, $2/1024}' To this one, replace example.net by the pool name of your site: ps --no-headers -e -o "pid,rss,cmd" | grep -E "[p]hp-fpm: pool example.net" | awk '{printf "PID %s: %.1f MB RSS\n", $1, $2/1024}' [image] perabyte: No, I ran that command and it came back empty. I looked at the journalctl for php8.3-fpm in general and I didn’t see any abnormal messages. If you don’t have such messages… could you please show the output of this command to check the limits used by nginx? curl -fsSLm30 https://7j.gg/chknof | sudo bash -s --

Fascinatingly, I just did a fail2ban-client unban --all as a sort of crapshoot attempt and the sites are magically alive again? I don’t know if this is just temporary though. Let me follow these steps just in case… [image] sahsanu: Show the output of these commands: Server one, first command: PID 5623: 67.2 MB RSS PID 6048: 62.0 MB RSS PID 6337: 63.1 MB RSS Server one, second command (for the site that was being dodgy): PID 6644: 67.8 MB RSS Server two, first command: PID 7508: 67.6 MB RSS PID 7512: 57.9 MB RSS PID 7513: 61.6 MB RSS Server two, second command (same situation): PID 8164: 56.5 MB RSS [image] sahsanu: If you don’t have such messages… could you please show the output of this command to check the limits used by nginx? First server: Checking services nginx apache2 The open files limit threshold has been set at 80% Process 1423 :: nginx: master process /usr/sbin/nginx -c /etc/nginx/nginx.conf Current open files: 88 Limit for open files: 1024 Process 1424 :: nginx: worker process Current open files: 114 Limit for open files: 65535 Process 1425 :: nginx: worker process Current open files: 90 Limit for open files: 65535 Process 1426 :: nginx: worker process Current open files: 90 Limit for open files: 65535 Process 1427 :: nginx: worker process Current open files: 91 Limit for open files: 65535 Process 1428 :: nginx: cache manager process Current open files: 87 Limit for open files: 65535 Second server: Checking services nginx apache2 The open files limit threshold has been set at 80% Process 1912 :: nginx: master process /usr/sbin/nginx -c /etc/nginx/nginx.conf Current open files: 158 Limit for open files: 1024 Process 1914 :: nginx: worker process Current open files: 170 Limit for open files: 65535 Process 1915 :: nginx: worker process Current open files: 160 Limit for open files: 65535 Process 1917 :: nginx: worker process Current open files: 160 Limit for open files: 65535 Process 1918 :: nginx: worker process Current open files: 160 Limit for open files: 65535 Process 1919 :: nginx: cache manager process Current open files: 155 Limit for open files: 65535 Thank you so much for your help walking me through this! I know this isn’t strictly a Hestia related issue, so I really appreciate it.

[image] perabyte: Fascinatingly, I just did a fail2ban-client unban --all as a sort of crapshoot attempt and the sites are magically alive again? Could you please send me the list of the ips so I can check them? awk '/Unban / {print $NF}' /var/log/fail2ban.log | sort -Vu I’ll check later your pool conf and limits.

[image] sahsanu: Could you please send me the list of the ips so I can check them? Here’s what I got: 1.235.192.214 2.57.122.177 2.57.122.238 4.209.219.178 4.213.160.153 5.2.124.162 5.9.120.241 5.35.248.158 5.39.70.39 5.44.111.80 5.44.111.149 8.219.1.122 14.63.217.28 20.153.204.5 23.92.78.16 23.227.167.26 27.128.160.208 31.47.253.137 34.14.122.221 36.66.16.233 36.83.226.90 36.95.194.50 37.27.141.218 37.58.136.133 37.97.185.98 38.137.11.14 38.242.139.95 39.96.6.171 39.105.216.234 40.81.31.179 41.128.181.199 41.226.29.18 42.96.20.16 42.200.66.164 43.128.60.40 43.130.33.245 43.133.61.254 43.162.103.6 43.164.195.160 43.245.249.251 45.12.28.218 45.61.184.228 45.148.10.183 46.16.90.98 46.163.72.164 46.243.95.178 47.76.100.35 47.94.4.180 47.97.118.250 49.232.24.93 50.6.229.90 51.75.64.35 51.77.158.34 51.91.124.115 51.195.62.42 54.38.241.200 54.242.108.234 58.147.171.11 59.22.201.143 62.27.5.128 62.108.32.112 62.108.32.138 62.148.185.5 62.210.122.54 64.188.77.26 69.5.23.222 74.208.42.164 74.208.81.179 75.119.150.220 75.119.208.70 77.237.158.44 81.88.34.74 81.88.36.57 81.169.140.37 81.169.215.92 82.39.86.153 82.165.189.108 82.220.38.103 82.223.120.80 83.168.95.29 83.177.240.110 83.224.175.43 84.247.133.95 85.5.148.125 85.25.46.229 85.214.28.25 85.214.43.7 85.214.47.63 85.214.56.101 85.214.60.164 85.214.91.4 85.214.133.120 85.214.147.216 85.214.154.50 85.214.158.155 85.214.203.242 85.214.208.71 85.214.237.67 85.215.78.221 85.215.200.66 85.215.215.169 85.234.143.175 85.236.38.172 87.95.167.222 87.106.44.172 87.106.53.29 87.106.120.204 87.118.94.212 87.140.48.96 87.226.190.225 87.229.120.48 89.58.12.180 91.92.40.5 91.92.40.8 91.92.40.11 91.92.40.13 91.92.40.204 92.51.135.140 92.204.37.150 92.204.58.4 92.204.239.201 92.205.50.42 92.205.53.59 92.205.54.48 92.205.54.144 92.205.54.254 92.205.161.152 92.205.165.63 92.205.185.113 92.205.193.232 92.205.196.153 92.205.208.129 92.205.211.106 92.205.225.165 93.47.117.83 94.46.164.66 94.102.213.67 94.177.162.41 95.110.227.50 95.173.103.13 101.47.15.119 101.47.27.73 101.231.133.210 102.88.137.80 102.220.85.131 103.5.210.47 103.75.182.132 103.101.162.56 103.103.245.61 103.148.100.146 103.149.176.34 103.183.62.2 103.189.208.13 103.237.144.204 103.243.24.124 106.38.195.164 107.170.40.174 107.172.13.216 109.71.253.107 109.235.63.171 112.124.51.174 113.177.27.200 114.111.54.189 116.177.174.231 118.26.36.195 118.178.187.12 118.193.38.26 119.209.12.20 119.246.15.94 120.48.2.29 123.10.98.105 125.76.228.194 130.51.180.75 130.180.91.158 132.145.213.106 133.18.122.63 134.119.86.47 138.113.22.151 138.197.130.251 139.59.4.137 139.59.208.49 139.99.38.177 146.120.164.13 147.135.220.202 147.182.183.153 148.72.159.0 159.195.51.249 159.223.213.49 160.25.166.88 162.19.73.19 162.215.13.26 162.215.130.111 163.5.143.4 163.7.6.41 164.92.161.148 165.99.16.135 165.227.159.13 167.99.133.242 167.99.148.102 167.172.208.8 171.25.158.68 171.25.158.87 172.104.5.198 172.190.89.127 173.231.240.237 173.236.193.42 175.11.107.92 178.104.220.36 178.175.167.68 178.254.6.141 179.228.54.137 181.218.172.74 182.52.90.106 185.3.235.166 185.18.198.75 185.23.70.17 185.44.132.5 185.55.229.76 185.103.143.101 185.118.198.210 185.201.28.88 185.201.254.82 185.219.40.250 185.243.11.57 185.243.11.127 185.243.11.129 185.243.11.161 186.96.158.180 187.110.238.50 188.68.47.105 188.68.47.223 190.129.122.12 193.43.145.120 194.124.73.166 195.178.106.224 195.178.110.30 197.5.145.102 198.11.178.150 199.247.30.79 202.61.232.193 202.61.232.223 202.61.233.72 202.145.0.61 203.56.46.78 203.142.160.143 207.180.199.91 207.180.219.250 209.159.149.206 212.83.57.172 212.132.78.209 212.227.94.44 213.136.73.80 213.165.87.111 213.209.159.158 217.73.30.34 217.76.55.220 217.79.182.225 217.154.70.17 217.154.239.16 217.160.90.172 217.182.193.175 217.196.145.108 218.144.90.40 218.157.205.238 222.179.105.58

[image] perabyte: Here’s what I got: I don’t see any IP belonging to Cloudflare: 20 Host Europe GmbH 16 IONOS SE 15 Strato GmbH 11 netcup GmbH 11 DigitalOcean LLC 10 OVH SAS 7 dogado GmbH 7 Contabo GmbH 6 Hangzhou Alibaba Advertising Co. Ltd. 5 Tencent Building Kejizhongyi Avenue 5 TechTies Inc. 5 Microsoft Corporation 5 Korea Telecom 4 BytePlus Pte. Ltd. 3 PT Telekomunikasi Indonesia 3 Hetzner Online GmbH 3 China Telecom 3 Alibaba (US) Technology Co. Ltd. 2 Vautron Rechenzentrum AG 2 Unmanaged Ltd 2 Unified Layer 2 UCloud Information Technology (HK) Limited 2 Techoff SRV Limited 2 Patrik Lagerman 2 Oracle Corporation 2 New Dream Network LLC 2 Meteverse Limited. 2 Interdigi Joint Stock Company 2 Hivelocity Inc. 2 Fastweb SpA 2 comtrance service GmbH 2 China Unicom China169 Backbone 2 China Telecom (Group) 2 Aruba S.p.A. 1 Wiit AG 1 Vodafone GmbH 1 VNPT Corp 1 VNET a.s. 1 velia.net Internetdienste GmbH 1 Velia.net 1 Uniserver Internet B.V. 1 TOT Public Company Limited 1 Total Play Telecomunicaciones S.A.P.I. de C.V. 1 TopNet 1 Timeweb LLP 1 TierPoint LLC 1 The Constant Company LLC 1 Telefonica Brasil S.A 1 Tele2 Sverige AB 1 Tedev Technological Development Company Limited 1 Team Blue Carrier Limited 1 Syrian Telecommunication Private Closed Joint Stock Company 1 Swisscom (Schweiz) AG 1 Studio AN-TV Srl 1 Stiegeler Internet Service GmbH 1 Speedbone Internet & Connectivity GmbH 1 Soluciones Web On Line S.L. 1 Societe Nationale Des Telecommunications (Tunisie Telecom) 1 Skypass Solutions Sp. Z.O.O. 1 SK Broadband Co Ltd 1 Signet B.V. 1 Shenzhen Tencent Computer Systems Company Limited 1 Server Plan S.r.l. 1 Scaleway SAS 1 RendszerNET Kft. 1 PT. Uninet Media Sakti (ISP) 1 PT Smart Media Pratama 1 PT Herza Digital Indonesia 1 PT. Garuda Prima Internetindo 1 PJSC Rostelecom 1 Oliver Horscht Is Trading As SYNLINQ 1 OC Networks Limited 1 NhanHoa Software Company 1 Netplus Broadband Services Private Limited 1 MTN Nigeria Communication Limited 1 Mobifone Corporation 1 Matteo Martelloni Trading As DELUXHOST 1 Long Van Soft Solution JSC 1 LINKdotNET 1 Keyweb AG 1 Kagoya Japan Inc. 1 IT Concept SRL 1 Interserver Inc 1 Internetx GmbH 1 InMotion Hosting Inc. 1 IDC China Telecommunications Corporation 1 HostPapa 1 Hong Kong Broadband Network Ltd 1 HKT Limited 1 HEXATOM s.a.r.l. 1 Hasan Broadband Net 1 Hanoi Telecom JSC 1 Google LLC 1 Gamma Telecom Holdings Ltd 1 Frantech Solutions 1 Feo Prest SRL 1 Fastpath Ike 1 Executive Service S.R.L. Societa Benefit 1 EVANZO e-commerce GmbH 1 Empresa Nacional de Telecomunicaciones Sociedad Anonima 1 Ecotel Communication AG 1 Dolphin IT-Systeme E.K. 1 DNA Oyj 1 Dianet Ltd. 1 Deutsche Telekom AG 1 Delta HighTech Ltd. 1 DB3 Servicos de Telecomunicacoes S.A 1 conova communications GmbH 1 CNServers LLC 1 Cloudie Limited 1 Cloud Computing Corporation 1 CloudAfrica Hosting (Pty) Ltd 1 Claro NXT Telecomunicacoes Ltda 1 CEZNET s.r.o. 1 BSE Software GmbH 1 Branch of Long Van System Solution JSC - Hanoi 1 Brainoza Ou 1 Beijing Baidu Netcom Science and Technology Co. Ltd. 1 Axarnet Comunicaciones S.L. 1 Association Librahost 1 Amazon.com Inc. 1 Almouroltec Servicos de Informatica e Internet Lda 1 Akamai Connected Cloud 1 Aftab IT Limited. 1 AdminVPS OOO 1 Adista SAS 1 ABN 1 23M GmbH So the timeouts are a bit strange, but too many firewall rules in the incorrect order can cause a lot of network problems. Show the output of this command (you can replace your public IP with 203.0.113.1): iptables -S | sed 's/here.your.public.ip/203.0.113.1/' Regarding the number of files limit, it’s ok, well nginx master process has only 1024 so it’s a bit low. Regarding php-fpm pool conf, I would use something like this: pm = dynamic pm.max_children = 80 pm.start_servers = 8 pm.min_spare_servers = 4 pm.max_spare_servers = 16 pm.max_requests = 500 But keep in mind that you have more sites, as I said it will depend on the free memory, the memory you want to assign to every site, etc. Above conf is for a normal site with some peeks of traffic. It will use at most this much RAM: 80 × 65 MB ≈ 5.2 GB For normal sites, the default conf is fine.

[image] sahsanu: Show the output of this command (you can replace your public IP with 203.0.113.1): Here you go, relatively similar results on both servers (looks like maybe fail2ban caught some IPs overnight) -P INPUT DROP -P FORWARD ACCEPT -P OUTPUT ACCEPT -N fail2ban-FTP -N fail2ban-HESTIA -N fail2ban-MAIL -N fail2ban-RECIDIVE -N fail2ban-SSH -N fail2ban-WEB -N hestia -A INPUT -p tcp -m multiport --dports 80,443 -j fail2ban-WEB -A INPUT -p tcp -m tcp --dport 8083 -j fail2ban-HESTIA -A INPUT -p tcp -m multiport --dports 25,465,587,110,995,143,993 -j fail2ban-MAI L -A INPUT -p tcp -m tcp --dport 21 -j fail2ban-FTP -A INPUT -p tcp -m tcp --dport 22 -j fail2ban-SSH -A INPUT -p tcp -m multiport --dports 1:65535 -j fail2ban-RECIDIVE -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT -A INPUT -s 10.1.0.5/32 -j ACCEPT -A INPUT -s 5.78.44.175/32 -j ACCEPT -A INPUT -s 127.0.0.1/32 -j ACCEPT -A INPUT -p tcp -m tcp --dport 22 -j ACCEPT -A INPUT -p tcp -m multiport --dports 80,443 -j ACCEPT -A INPUT -p tcp -m multiport --dports 21,12000:12100 -j ACCEPT -A INPUT -p udp -m udp --dport 53 -j ACCEPT -A INPUT -p tcp -m tcp --dport 53 -j ACCEPT -A INPUT -p tcp -m multiport --dports 25,465,587 -j ACCEPT -A INPUT -p tcp -m multiport --dports 110,995 -j ACCEPT -A INPUT -p tcp -m multiport --dports 143,993 -j ACCEPT -A INPUT -p tcp -m tcp --dport 8083 -j ACCEPT -A INPUT -p icmp -j ACCEPT -A fail2ban-FTP -j RETURN -A fail2ban-HESTIA -j RETURN -A fail2ban-MAIL -j RETURN -A fail2ban-RECIDIVE -s 2.57.122.177/32 -j REJECT --reject-with icmp-port-unreac hable -A fail2ban-RECIDIVE -s 195.178.110.30/32 -j REJECT --reject-with icmp-port-unre achable -A fail2ban-RECIDIVE -j RETURN -A fail2ban-SSH -j RETURN -A fail2ban-WEB -j RETURN -P INPUT DROP -P FORWARD ACCEPT -P OUTPUT ACCEPT -N fail2ban-FTP -N fail2ban-HESTIA -N fail2ban-MAIL -N fail2ban-RECIDIVE -N fail2ban-SSH -N fail2ban-WEB -N hestia -A INPUT -p tcp -m multiport --dports 80,443 -j fail2ban-WEB -A INPUT -p tcp -m tcp --dport 8083 -j fail2ban-HESTIA -A INPUT -p tcp -m multiport --dports 25,465,587,110,995,143,993 -j fail2ban-MAI L -A INPUT -p tcp -m tcp --dport 21 -j fail2ban-FTP -A INPUT -p tcp -m tcp --dport 22 -j fail2ban-SSH -A INPUT -p tcp -m multiport --dports 1:65535 -j fail2ban-RECIDIVE -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT -A INPUT -s 5.78.142.129/32 -j ACCEPT -A INPUT -s 127.0.0.1/32 -j ACCEPT -A INPUT -s 10.1.0.3/32 -p tcp -j ACCEPT -A INPUT -p tcp -m tcp --dport 22 -j ACCEPT -A INPUT -p tcp -m multiport --dports 80,443 -j ACCEPT -A INPUT -p tcp -m multiport --dports 21,12000:12100 -j ACCEPT -A INPUT -p udp -m udp --dport 53 -j ACCEPT -A INPUT -p tcp -m tcp --dport 53 -j ACCEPT -A INPUT -p tcp -m multiport --dports 25,465,587 -j ACCEPT -A INPUT -p tcp -m multiport --dports 110,995 -j ACCEPT -A INPUT -p tcp -m multiport --dports 143,993 -j ACCEPT -A INPUT -p tcp -m tcp --dport 8083 -j ACCEPT -A INPUT -p icmp -j ACCEPT -A fail2ban-FTP -j RETURN -A fail2ban-HESTIA -j RETURN -A fail2ban-MAIL -j RETURN -A fail2ban-RECIDIVE -j RETURN -A fail2ban-SSH -j RETURN -A fail2ban-WEB -j RETURN [image] sahsanu: Regarding the number of files limit, it’s ok, well nginx master process has only 1024 so it’s a bit low. Would you suggest raising it, and do you know how I’d go about that? I’m not sure what a good number would be. [image] sahsanu: Regarding php-fpm pool conf, I would use something like this: Will try this out for the sites with above average traffic! Thank you!!

Regarding the firewall, Hestia doesn’t add the rules efficiently, instead of: -A INPUT -p tcp -m multiport --dports 80,443 -j fail2ban-WEB -A INPUT -p tcp -m tcp --dport 8083 -j fail2ban-HESTIA -A INPUT -p tcp -m multiport --dports 25,465,587,110,995,143,993 -j fail2ban-MAI L -A INPUT -p tcp -m tcp --dport 21 -j fail2ban-FTP -A INPUT -p tcp -m tcp --dport 22 -j fail2ban-SSH -A INPUT -p tcp -m multiport --dports 1:65535 -j fail2ban-RECIDIVE -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT -A INPUT -s 10.1.0.5/32 -j ACCEPT -A INPUT -s 5.78.44.175/32 -j ACCEPT -A INPUT -s 127.0.0.1/32 -j ACCEPT It should look like this: -A INPUT -s 127.0.0.1/32 -j ACCEPT -A INPUT -s 10.1.0.5/32 -j ACCEPT -A INPUT -s 5.78.44.175/32 -j ACCEPT -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT -A INPUT -p tcp -m multiport --dports 80,443 -j fail2ban-WEB -A INPUT -p tcp -m tcp --dport 8083 -j fail2ban-HESTIA -A INPUT -p tcp -m multiport --dports 25,465,587,110,995,143,993 -j fail2ban-MAI L -A INPUT -p tcp -m tcp --dport 21 -j fail2ban-FTP -A INPUT -p tcp -m tcp --dport 22 -j fail2ban-SSH -A INPUT -p tcp -m multiport --dports 1:65535 -j fail2ban-RECIDIVE When I have some free time, I’ll check it and open a PR to modify the v-update-firewall script. [image] perabyte: Would you suggest raising it, and do you know how I’d go about that? 65536 is a good number. To modify it for nginx: mkdir -p /etc/systemd/system/nginx.service.d/ echo -e '[Service]\nLimitNOFILE=65536' > /etc/systemd/system/nginx.service.d/override.conf systemctl daemon-reload systemctl restart nginx You can checkt it again and instead of 1024 you will see 65536: curl -fsSLm30 https://7j.gg/chknof | sudo bash -s --

[image] sahsanu: When I have some free time, I’ll check it and open a PR to modify the v-update-firewall script. That’s very good to know… perhaps iptables was the source of my woes. I’ll keep an eye out for that update! [image] sahsanu: 65536 is a good number. I have made the change! Thank you again for all your help! I learned a lot from this thread. :blush:

Timeouts on Specific Websites

Community Support Web

sahsanu June 18, 2026, 5:04pm 2

Hi,

Are you sure? Take a look to this topic just in case:

It looks like those sites are exhausting the pm workers.

What did you try and what is the current php-fpm conf for those sites?

Did you see any max_children message in the logs?

journalctl -u php8.3-fpm | grep max_children