Docker sysdig installation error:

According to the online https://www.cnblogs.com/lkun/p/7898900.html centos7 install docker monitoring – sysdig (solve installation errors) did not successfully solved
 
Then I saw this reply under Cloudman’s message:

 
1, his host host kernel upgrade, upgrade instructions: yum update – y kernel, using the uname -r see if upgrade success

2. Install yum -y kernel-devel-$(uname -r) to install the kernel header file in the host operating system
Sysdig command: sysdig command

docker container run -it --rm --name=sysdig --privileged=true  --volume=/var/run/docker.sock:/host/var/run/docker.sock  --volume=/dev:/host/dev  --volume=/proc:/host/proc:ro  --volume=/boot:/host/boot:ro  --volume=/lib/modules:/host/lib/modules:ro  --volume=/usr:/host/usr:ro  sysdig/sysdig

Input csysdig:

 

>

docker rmi $(docker images | grep "none" | awk '{print $3}')

:
ror response from daemon: conflict: Unable to delete 8598c91556DC (must be forced) – image is being used by stopped container 8f10ef9b804f
Because mirroring is used by a container, you need to delete the container first, and then execute the delete command

docker rm 8f10ef9b804f

docker rmi $(docker images | grep "none" | awk '{print $3}')

Docker 19.03 for July 2019 has been officially released, and for me there are two highlights of this release.
1, is the docker don’t need root access to start drinking run
2, is to support the GPU enhancements, we want to read in the docker nvidia graphics card no longer need extra install nvidia – docker
Verify GPU options:
docker run –help | grep -i gpus
Docker calls the GPU command:
docker run -it –rm –gpus all ubuntu /bin/bash
The training model
parameter — GPUS ID must be present, otherwise the GPU will not be used. ID refers to which GPU is used for training, and it refers to binding the container to the specified GPU for running
 
Refer to the blog: https://blog.csdn.net/byaa34616/article/details/100951724

Solution:
systemctl stop firewalld.
systemctl disable firewalld.
s>mctl disable firewalld.
systemctl disable firewalld.
systemctl disable firewalld.
systemctl disable firewalld.
systemctl disable firewalld.
systemctl disable firewalld If Tomcat is still inaccessible after the firewall is turned off, the webapps folder is empty
2. The problem of webapps folder is empty check the tomcat container ID

docker ps
enter tomcat container terminal
docker exec – it tomcat container ID/bin/bash
enter
CD webapps webapps folder
copy on a webapps directory. All files to the webapps directory under the dist
cp – r.. /webapps.dist/* ./

1. Dockerfile build file defines parameters

FROM java:8
VOLUME ["/home/pdsstm/uploads/pdsfile"]
COPY pds-stm-management.jar /home/pds-stm/jar/pds-stm-management.jar
ENV CE=$CE
EXPOSE 31080
ENTRYPOINT [ "sh", "-c", "java  -Djava.security.egd=file:/dev/./urandom -jar /home/pds-stm/jar/pds-stm-management.jar --spring.profiles.active=$CE" ]

Define variable CE
2. Go to the folder to build the image

docker build -t pds-stm .

3. Run the image with parameters

docker run -d -p 38082:31080 --name pds-stm -e CE="test" -v /home/pds-stm/jar:/home/pds-stm/jar -v /home/pds-stm/logs:/logs  --restart=always pds-stm

 

FAQ: 1, Cannot connect to the Docker daemon. Is the Docker daemon running on this host?Solution: This is the most common problem with docker when it’s first installed. There are two standard ways to solve this problem: A, start the Docker background service with root authority, and Service Docker Start. B, or add the current user to the Docker group. Apart from these two reasons, Docker daemon could not start, but more because the system boot time was wrong. Because the Gentoo kernel is self-defined, the majority of error reporting may be caused by parameter configuration in the kernel. Check /var/log/docker.log for any errors.
Error starting Daemon: Devices CGroup isn’t mounted in the kernel of Gentoo. It is required to open the cGroup, such as memory CFS RT, etc. For the BLkio error of cGroup, it needs to be turned on through the Settings inside the block.
3, Can’t Initialize iptables table ‘NAT ‘: Table does not Exist solution: Because iptables is the dependent component required to run for Docker (security Settings are required), it is crucial to set up the iptable correctly. If NAT is wrong when Docker starts, there are usually two reasons. The first reason is that iptables is not set correctly, and the second reason is the kernel setting. The standard iptables setup sequence is as follows: Iptables -F Service IPtables Save Service Iptables Restart has Gentoo installed and configured properly and should theoretically have an IPtable and NAT side. If the IPtables or NAT Settings are not correct, there is a problem with the kernel configuration. In kernel configuration, to activate IPtable and NAT, it is necessary to open netFilter and all the following functions in the network. Otherwise, at runtime, you’ll get all sorts of weird bugs. After the compilation core is restarted, the problem is resolved.
4, Error starting daemon: error initializing graphdriver: devmapper: Base Device UUID and F ilesystem verification failed: devicemapper: Error running deviceCreate (ActivateDevice) dm_task_run failed, fault analysis, the causes of the failure of several, one is the docker storage space is full, the new add or modify the container will be an Error, another reason is more direct,/var/log/docker/under the devicemapper data and metadata file is deleted. But the bottom line is that the uUID of the existing Storage-set is inconsistent with the UUID saved in the Docker configuration. Solution: The solution to this problem is simple and brutal. Clean all the configuration files of Docker directly (all the existing data will be lost), restart the service, and let Docker rebuild automatically. rm -rf /var/lib/docker/* service docker restart
 
Copyright belongs to the author. Please contact the author for any reprint.
the author: keenshoes (from douban)
source: https://www.douban.com/note/636127425/