Open captainabloc opened 6 months ago
Is it the index.js process or nginx?
Also do you use crowdsec or have you enabled modsec on any host?
hi, it's nginx process - up to 100% CPU when editing advances I use crowdsec
Can you disable crowdsec and try again?
got some modsecurity enabled in other proxies, all removed and now much faster....
Yes, modsecurity often memory leak or uses a lot cpu when reloading, same for crowdsec sometimes. There is nothing I can do to prevent this. Either you accept this high usage or don't use modsec... I know not a good solution, but there is nothing I can do
I've pinned this
Hi,
So this was probably caused by crowdsec, lua has a bad memory leak if you have use lua_ssl_trusted_certificate
since this gets included into each and every location block.
I fixed this in my npm fork by forcing just googles cert instead of the system keyring.
lua_ssl_trusted_certificate /etc/ssl/certs/GTS_Root_R1.pem
I've been having this problem too, but the problem is ModSecurity or CoreRuleSet? I think I'm not using crowdsec (I don't even know how to enable or disable it).
Asking because I remember that when I went on to disable ModSecurity and CoreRuleSet for all the proxy hosts, when the only ones that were left were the ones that only had ModSecurity enabled the high CPU consumption stopped, I still disabled ModSecurity in everything just in case because the problem made the entire system halt but I think ModSecurity is not triggering this issue.
coreruleset has a lot of rules, this can cause hugh memory usage if enabled on many hosts, sorry, but that is "normal" behavior
Hi,
So this was probably caused by crowdsec, lua has a bad memory leak if you have use
lua_ssl_trusted_certificate
since this gets included into each and every location block.I fixed this in my npm fork by forcing just googles cert instead of the system keyring.
lua_ssl_trusted_certificate /etc/ssl/certs/GTS_Root_R1.pem
I have the same problem, I only use crowdsec, modesecurity is not used.
Docker with NPMplus runs on a Proxmox container. Even with 8 GB ram, the workload is 100% after a few days. The proxy then no longer works.
I would like to try your solution, can you tell me where I have to make the adjustment?
Hello, all works fine, but RAm consumption is constantly growing and is now up to 2 Gb alone, what seems high on a linux system. Also any edition of advanced settings brings CPU consumption to 100% for a minute or more.
Docker runs on a dedicated debian 12 VM on Proxmox, with
would anyone know how to limit that?
many thanks