Ruby 1.8.7 on Rails 2.3.5 application rendering speed slows down overtime

Previously, the application worked on a shared hosting without a docker container and the performance was good. After transferring to a dedicated server, the application was made through a docker container and a performance problem appeared. The developer assured that there was a problem with the old Ruby, but if before everything worked within the normal range, now the application restarts every 6 hours trough cron (docker-compose down) to maintain normal speed. Application handles concurrency poorly.

The main problem is that initial rendering speed is:

Processing CatalogController#index (for 127.0.0.1 at 2023-09-06 19:22:06) [GET] Parameters: {"controller"=>"catalog", "language"=>"ru", "action"=>"index"} Rendering template within layouts/main Rendering catalog/index Completed in 330ms (View: 300, DB: 30) | 200 OK [http://localhost/]

Than after 300 requests it slows to:

Processing CatalogController#index (for 127.0.0.1 at 2023-09-06 19:22:06) [GET] Parameters: {"controller"=>"catalog", "language"=>"ru", "action"=>"index"} Rendering template within layouts/main Rendering catalog/index Completed in 1266ms (View: 1233, DB: 37) | 200 OK [http://localhost/]

When I try to use ab -n 100 http://localhost:3000/ it shows result like:

Concurrency Level:      1
Time taken for tests:   33.883 seconds
Complete requests:      100
Failed requests:        0
Total transferred:      6796500 bytes
HTML transferred:       6764800 bytes
Requests per second:    2.95 [#/sec] (mean)
Time per request:       338.831 [ms] (mean)
Time per request:       338.831 [ms] (mean, across all concurrent requests)
Transfer rate:          195.89 [Kbytes/sec] received

Code in Dockerfile that builds application:

FROM centos:7
SHELL ["/bin/bash", "-c"] 

RUN yum update -y && yum -y install gpg curl which wget ImageMagick-devel mysql mysql-devel libxml2-devel libxslt libxslt-devel

#Install RVM
RUN gpg --keyserver keyserver.ubuntu.com --recv-keys 409B6B1796C275462A1703113804BB82D39DC0E3 7D2BAF1CF37B13E2069D6956105BD0E739499BDB
RUN \curl -sSL https://get.rvm.io | bash -s stable

#Install Ruby
RUN source /etc/profile.d/rvm.sh && rvm install 1.8.7

#Install Ruby Gems
RUN source /etc/profile.d/rvm.sh && rvm use 1.8.7 && gem install -v=2.3.5 rails && gem install bundler -v=1.3.5 && gem install rubygems-update -v 1.3.7 && update_rubygems --version=1.3.7 && gem install mime-types -v=1.17.2 && gem install russian -v=0.2.7 && gem install httparty -v=0.7.2 && gem install mini_mime -v=1.1.2 && gem install yandex_inflect -v=0.1.0 && PATH="/app/ImageMagick-6.8.0-10/Magick++/bin:$PATH" gem install rmagick -v=2.13.2 && gem install mechanize -v=0.8.5 && gem install geoip -v 1.1.2 && gem install rdoc -v=3.9.2 && gem install mysql -v=2.9.1 && gem install nokogiri -v=1.4.4 && gem install daemons -v=1.1.0 && gem install thin -v=1.7.0

WORKDIR /app

RUN mkdir /app/tmp

CMD source /etc/profile.d/rvm.sh && rvm use 1.8.7 && script/server thin -p 3000 -e production

Maybe someone knows what can cause such a behavior?

I have tried to install thin server instead of a WebBrick but it didn't resolve the problem. And only solution for now is docker-compose down command but it restarts application entirely for almost 16 seconds. Lookes through a code and didn't find code that can slow down the application.

Possible solutions to the problem:

1. Correctly configure the docker container (the application did not work on docker before, so the problem may be in its settings)
2. Install unicorn or passenger (it worked on webrick during the transfer, then transferred to thin, but it didn't help much)
3. Install some additional gems (before, more gems were installed on the hosting, but now only the most necessary ones).
4. It is necessary to transfer the application to a new version of Ruby on Rails (if there is no other way out, then you have to do it)