폐쇄망환경에 구성된 kubernetes 상에서 http(s)_proxy 설정을 하는 방법에 대하여 간단히 알아보자. 사전에 proxy 서버를 구성하고 해당 주소는 다음과 같다고 가정하고 다음 설정을 진행한다. proxy 서버 주소 : https://192.168.1.1:3128 사용하려는 pod 혹은 deployment 에 다음 항목을 env 위치에 추가한다. (deployment나 replicaset에 추가해주는것이 향후 재사용을 위해서도 좋다.) spec: template: spec: containers: ... envs: - name: https_proxy value: http://192.168.1.1:3128 - name: http_proxy value: http://192.168.1.1:3128..
일반적으로 사용자의 요청에 대한 error 가 발생되면 origin 서버에서 error page를 rendering 하여 응답하게 된다. 만약 proxy 환경이라면 이를 proxy 서버에서 처리할수 있게 설정이 가능하다. location ~ \.(7z|avi|avif|apkbin|bmp|bz2|png|...|jpg)$ { ... proxy_pass http://192.168.1.1; proxy_intercept_errors on; error_page 404 500 501 502 503 /error_page.html; ... } location = /error_page.html { root /tmp; ##
먼저 가정은 별도의 lua module loading 과정은 사전에 이루어져 있다 보고 기술하였다. 기본 workflow를 알아둘 필요가 있다. https://cloud.githubusercontent.com/assets/2137369/15272097/77d1c09e-1a37-11e6-97ef-d9767035fc3e.png 위 flow에 기반하여 lua 모듈이 동작되어질 위치를 확인하고 다음과 같은 module을 통해 호출을 하면된다. (주로 사용해본 모듈은 rewrite_by_lua_xxx, access_by_lua_xxx, content_by_lua_xxx, set_by_lua_xxx 등이 있다.) https://github.com/openresty/lua-nginx-module#rewrite_by_..
현상 아래와 같은 message log 출력과 함께 ssh 연결이 간헐적으로 연결되지 않는 현상발생 Feb 14 17:30:07 ubuntu-instance kernel: net_ratelimit: 27 callbacks suppressed Feb 14 17:30:33 ubuntu-instance kernel: net_ratelimit: 10 callbacks suppressed 혹은 syslog에 다음과 같은 메세지가 출력될수도 있다. Neighbour table overflow 원인 다수의 arp table이 생기는 가상화 환경에서 발생될수 있는 문제로 기본적으로 아래와 같은 gc_thresh value를 가지고 있어 그 이상이 생길경우 문제가 발생된다. root@ubuntu-instance:/hom..
nginx-ingress를 사용하면서 ingress에 nginx 설정을 좀더 추가하고 싶고 때에 따라서는 nginx-ingress에 글로벌하게 설정이 필요한 경우도 있다. 하여 이를 설정하는 방법에 대하여 알아보도록 하겠다. 여기서는 log_format을 custom한 설정추가의 하나의 예로 기술하였다. 기본적인 custom 설정 추가 먼저 기본적으로 custom 설정을 추가하기 위한 기본 작업을 알아보면, nginx-ingress deployment를 확인해보면 다음과 같은 container 실행시 추가될 argument가 정의되어 있다. ... spec: containers: - args: - /nginx-ingress-controller - --election-id=ingress-controller..
이슈사항 setuptools 의 버전 문제로 현재 유지보수가 이루어지지 않는 python 2.7을 사용하려 할때 다음과 같은 syntax 에러를 맞이하는 경우가 있다. [vagrant@localhost ~]$ pip install --upgrade pip Collecting pip Using cached https://files.pythonhosted.org/packages/da/f6/c83229dcc3635cdeb51874184241a9508ada15d8baa337a41093fab58011/pip-21.3.1.tar.gz Complete output from command python setup.py egg_info: Traceback (most recent call last): File "", lin..
sockperf란 성능 테스팅을 위해 설계된 socket API를 기반으로 한 네트워크 벤치마킹 도구이다. sockperf build 별도의 binary 파일을 제공하고 있지 않기에 ubuntu 기준으로 다음과 같이 build를 하여 sockperf binary를 만들어 사용한다. sudo apt install perl make automake autoconf m4 libtool-bin g++ ./autogen.sh (only when cloning from repository) ./configure make && make install 빌드 당시 사용했던 OS 버전은 ubuntu 20.04 이다. sockperf 사용 서버/클라이언트 구조로 동작된다. 먼저 sockperf의 help 를 확인해보면 ub..
/etc/security/limits.conf 내에 nofile 을 변경했음에도 특정 프로세스는 적용이 안된것을 보게되어 이에 대해 확인한 사항을 정리한다. 이슈 확인 아래와 같이 ulimit으로 확인한 max open files 가 65535임에도 [root@localhost ~]# ulimit -n 65535 systemd에 의해 실행되는 process의 limits은 다른 값을 출력하는것을 확인하였다. [root@localhost ~]# cat /proc/11223/limits | grep "open file" Max open files 1024 4096 files 분석결과 systemd는 기본적으로 기본값을 따로 가지고 있고 해당 값이 적용되는것으로 보인다. $ sudo systemctl show..
- Total
- Today
- Yesterday
- azure policy
- kubernetes
- crashloopbackoff
- aquasecurity
- kata container
- Jenkinsfile
- mattermost
- ceph
- K3S
- metallb
- openstack backup
- Terraform
- nginx-ingress
- OpenStack
- boundary ssh
- minio
- hashicorp boundary
- kubernetes install
- jenkins
- DevSecOps
- minikube
- vmware openstack
- socket
- GateKeeper
- wsl2
- macvlan
- ansible
- open policy agent
- Helm Chart
- openstacksdk
일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | |||
5 | 6 | 7 | 8 | 9 | 10 | 11 |
12 | 13 | 14 | 15 | 16 | 17 | 18 |
19 | 20 | 21 | 22 | 23 | 24 | 25 |
26 | 27 | 28 | 29 | 30 | 31 |