Chore: Fixed Docker Health Check

[manojks1999]

Fixes #2650

[romange]

hello @manojks1999 . are you sure it's the right filter?. Have you tried running a docker image and run this command?

[danriedl]

I guess this won't work on the default container image.

dri@srv1> docker run -it --rm ghcr.io/dragonflydb/dragonfly sh -c "netstat -tuln | grep -oE ':::[0-9]+' | grep -oE '[0-9]+'"
sh: 1: netstat: not found

[romange]

Just confirmed, this fix is incorrect. @danriedl it is possible to exec as root to the already existing container:

docker exec -it -u 0 dfserver /usr/bin/bash and then install net-tools and run the command.

[manojks1999]

Hi @romange , made a change in the regular expression of the grep command, and I have made a test on the alpine docker image I have shared the logs here, and keep this on hold I will make a test on ubuntu image.

Testing this is delayed as building the docker image is taking time. I will update here once with proper testing in the Ubuntu image.

{
        "Id": "457e9699eb13317d840957027dac99cc2392b53c272d2fd3eb7a63383d1beb52",
        "Created": "2024-02-28T19:50:15.507886048Z",
        "Path": "entrypoint.sh",
        "Args": [
            "--port",
            "6767"
        ],
        "State": {
            "Status": "running",
            "Running": true,
            "Paused": false,
            "Restarting": false,
            "OOMKilled": false,
            "Dead": false,
            "Pid": 54519,
            "ExitCode": 0,
            "Error": "",
            "StartedAt": "2024-02-28T19:50:18.506419807Z",
            "FinishedAt": "0001-01-01T00:00:00Z",
            "Health": {
                "Status": "healthy",
                "FailingStreak": 0,
                "Log": [
                    {
                        "Start": "2024-02-29T01:21:51.124907738+05:30",
                        "End": "2024-02-29T01:21:52.352651502+05:30",
                        "ExitCode": 0,
                        "Output": "+PONG\r\n"
                    },
                    {
                        "Start": "2024-02-29T01:22:22.356281589+05:30",
                        "End": "2024-02-29T01:22:23.661224958+05:30",
                        "ExitCode": 0,
                        "Output": "+PONG\r\n"
                    },
                    {
                        "Start": "2024-02-29T01:22:53.662909485+05:30",
                        "End": "2024-02-29T01:22:54.954001121+05:30",
                        "ExitCode": 0,
                        "Output": "+PONG\r\n"
                    },
                    {
                        "Start": "2024-02-29T01:23:24.95819494+05:30",
                        "End": "2024-02-29T01:23:26.327866369+05:30",
                        "ExitCode": 0,
                        "Output": "+PONG\r\n"
                    },
                    {
                        "Start": "2024-02-29T01:23:56.329411658+05:30",
                        "End": "2024-02-29T01:23:57.615890188+05:30",
                        "ExitCode": 0,
                        "Output": "+PONG\r\n"
                    }
                ]
            }
        }
}

[manojks1999]

Hi @romange below is the inspect log of ubuntu docker file, it seems working fine,

{
        "Id": "614ad12239194bb2cee2f22063b15eb85e4aa75604eefb596f78d57fc4db038c",
        "Created": "2024-02-29T02:51:27.781317762Z",
        "Path": "entrypoint.sh",
        "Args": [
            "--port",
            "6869"
        ],
        "State": {
            "Status": "running",
            "Running": true,
            "Paused": false,
            "Restarting": false,
            "OOMKilled": false,
            "Dead": false,
            "Pid": 24812,
            "ExitCode": 0,
            "Error": "",
            "StartedAt": "2024-02-29T02:51:28.912175782Z",
            "FinishedAt": "0001-01-01T00:00:00Z",
            "Health": {
                "Status": "healthy",
                "FailingStreak": 0,
                "Log": [
                    {
                        "Start": "2024-02-29T08:21:58.920502738+05:30",
                        "End": "2024-02-29T08:22:00.061591505+05:30",
                        "ExitCode": 0,
                        "Output": "+PONG\r\n"
                    }
                ]
            }
        }
}

[manojks1999]

I just now found out this fails when we provide more than one argumental port.

docker run ubuntu-testing --port 5676 --memcached_port 4323

netstat -tul
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State      
tcp        0      0 0.0.0.0:4323            0.0.0.0:*               LISTEN     
tcp        0      0 0.0.0.0:5676            0.0.0.0:*               LISTEN     

do we want here to only monitor the port 5676 as provided and also to monitor the other ports that are exposed?

If we want only to monitor server port 5676 can I know how exactly we are accessing this docker argument in the code?

[romange]

cat /proc/1/cmdline will get you the command line with all the arguments.
But I think that it may be simpler to limit healthcheck heuristic to find a custom port the way you do now, but also
support the environment variable HEALTHCHECK_PORT, that in case it is provided to container via -e, the healthcheck uses its value for the port to connect.
so in the example above one would need to pass: docker run -e HEALTHCHECK_PORT=5676 ubuntu-testing --port 5676 --memcached_port 4323

[manojks1999]

Hi @romange ,

Updated the health check file accordingly, and I have tested it is working as expected. If this is approved, later we need to change the documentation with the option as passing environment variable for health check.

[romange]

Thank you for fixing this issue!

[romange]

@manojks1999 Thank you for fixing the issue! Added a comment - since for me locally (not in docker) the grep line did not work, s just double checking you tested the fix.