Amazon-ssm-agent unrecognized service (just installed it via Docker) - linux

Question

I am trying to figure out why I cannot start and stop the amazon-ssm-agent service manually in a Kali Linux Focker image running on an Ubuntu 20.04.1 LTS host. Per their instructions, I have obtained the .deb file and installed it with dpkg -i. Although I can interact with it via amazon-ssm-agent -h and registering it just fine, etc., I cannot restart the service which sometimes fixes the Connection Lost issue after registering.
As you can see below, I am using wget to get the .deb file, and installing it:
➜ ~ wget https://s3.us-east-1.amazonaws.com/amazon-ssm-us-east-1/latest/debian_amd64/amazon-ssm-agent.deb
--2020-12-27 22:21:32-- https://s3.us-east-1.amazonaws.com/amazon-ssm-us-east-1/latest/debian_amd64/amazon-ssm-agent.deb
Resolving s3.us-east-1.amazonaws.com (s3.us-east-1.amazonaws.com)... 52.217.109.126
Connecting to s3.us-east-1.amazonaws.com (s3.us-east-1.amazonaws.com)|52.217.109.126|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 41537900 (40M) [binary/octet-stream]
Saving to: 'amazon-ssm-agent.deb'
amazon-ssm-agent.deb 100%[========================================================================================================================================================================================================================================>] 39.61M 105MB/s in 0.4s
2020-12-27 22:21:33 (105 MB/s) - 'amazon-ssm-agent.deb' saved [41537900/41537900]
➜ ~ dpkg -i amazon-ssm-agent.deb
Selecting previously unselected package amazon-ssm-agent.
(Reading database ... 231292 files and directories currently installed.)
Preparing to unpack amazon-ssm-agent.deb ...
Preparing for install
Unpacking amazon-ssm-agent (3.0.431.0-1) ...
Setting up amazon-ssm-agent (3.0.431.0-1) ...
Starting agent
➜ ~ service amazon-ssm-agent status
amazon-ssm-agent: unrecognized service
➜ ~
I also cannot use systemctl because of the following error:
➜ ~ systemctl status amazon-ssm-agent
System has not been booted with systemd as init system (PID 1). Can't operate.
Failed to connect to bus: Host is down
➜ ~
I tried looking in /etc/init.d as well, but no luck:
➜ ~ ls /etc/init.d -l
total 240
-rwxr-xr-x 1 root root 2489 Aug 8 07:47 apache-htcacheclean
-rwxr-xr-x 1 root root 8181 Aug 8 07:47 apache2
-rwxr-xr-x 1 root root 1614 Jul 14 2019 atftpd
-rwxr-xr-x 1 root root 2401 May 26 2020 avahi-daemon
-rwxr-xr-x 1 root root 1175 Apr 17 2020 binfmt-support
-rwxr-xr-x 1 root root 2948 Sep 16 07:49 bluetooth
-rwxr-xr-x 1 root root 1232 Dec 1 01:02 console-setup.sh
-rwxr-xr-x 1 root root 937 Sep 3 22:30 cryptdisks
-rwxr-xr-x 1 root root 896 Sep 3 22:30 cryptdisks-early
-rwxr-xr-x 1 root root 3152 Jul 2 13:19 dbus
-rwxr-xr-x 1 root root 1408 Aug 4 23:00 dns2tcp
-rwxr-xr-x 1 root root 7159 May 23 2020 exim4
-rwxr-xr-x 1 root root 3708 Nov 25 21:07 hwclock.sh
-rwxr-xr-x 1 root root 3615 Sep 5 2019 inetsim
-rwxr-xr-x 1 root root 4113 Sep 26 16:48 iodined
-rwxr-xr-x 1 root root 1479 Oct 9 2016 keyboard-setup.sh
-rwxr-xr-x 1 root root 2044 Apr 18 2020 kmod
-rwxr-xr-x 1 root root 5966 Nov 22 15:42 mariadb
-rwxr-xr-x 1 root root 2882 Jul 26 2019 miredo
-rwxr-xr-x 1 root root 4486 Sep 21 14:45 networking
-rwxr-xr-x 1 root root 5658 Jul 26 12:02 nfs-common
-rwxr-xr-x 1 root root 4579 May 28 2020 nginx
-rwxr-xr-x 1 root root 1934 Jul 7 05:55 nmbd
-rwxr-xr-x 1 root root 1494 Sep 23 11:46 ntp
-rwxr-xr-x 1 root root 9138 Oct 28 18:37 openvpn
-rwxr-xr-x 1 root root 3720 Jun 14 2020 pcscd
-rwxr-xr-x 1 root root 1490 Nov 15 2019 postgresql
-rwxr-xr-x 1 root root 924 May 16 2020 procps
-rwxr-xr-x 1 root root 3699 Jul 22 2017 ptunnel
-rwxr-xr-x 1 root root 3836 Jan 2 2017 redsocks
-rwxr-xr-x 1 root root 1615 Aug 19 2018 rlinetd
-rwxr-xr-x 1 root root 2507 Jul 13 01:22 rpcbind
-rwxr-xr-x 1 root root 4417 Aug 26 20:23 rsync
-rwxr-xr-x 1 root root 2864 Oct 20 19:45 rsyslog
-rwxr-xr-x 1 root root 1661 Jun 5 2013 rwhod
-rwxr-xr-x 1 root root 2259 Jul 7 05:55 samba-ad-dc
-rwxr-xr-x 1 root root 1222 Apr 2 2017 screen-cleanup
-rwxr-xr-x 1 root root 3088 Oct 10 2019 smartmontools
-rwxr-xr-x 1 root root 2061 Jul 7 05:55 smbd
-rwxr-xr-x 1 root root 1175 Sep 24 23:10 snmpd
-rwxr-xr-x 1 root root 4056 Dec 2 10:32 ssh
-rwxr-xr-x 1 root root 4440 Sep 5 2019 sslh
-rwxr-xr-x 1 root root 5730 Sep 13 10:43 stunnel4
-rwxr-xr-x 1 root root 1030 Dec 2 03:10 sudo
-rwxr-xr-x 1 root root 1581 Dec 16 08:36 sysstat
-rwxr-xr-x 1 root root 6871 Dec 3 22:53 udev
-rwxr-xr-x 1 root root 2757 Oct 9 08:13 x11-common
➜ ~
However, you can see that running the amazon-ssm-agent command works just fine:
➜ ~ amazon-ssm-agent
Error occurred fetching the seelog config file path: open /etc/amazon/ssm/seelog.xml: no such file or directory
Initializing new seelog logger
New Seelog Logger Creation Complete
2020-12-27 22:24:08 ERROR error fetching the instanceID, Failed to fetch instance ID. Data from vault is empty. EC2MetadataError: failed to make EC2Metadata request
status code: 404, request id:
caused by: not found
2020-12-27 22:24:08 ERROR error occurred when starting amazon-ssm-agent: error fetching the instanceID, Failed to fetch instance ID. Data from vault is empty. EC2MetadataError: failed to make EC2Metadata request
status code: 404, request id:
caused by: not found
➜ ~
The only reason that I need to restart the service after registering is because sometimes I get a "Connection Lost" on the managed instance's ping status after registering. Usually restarting the service seem to do the trick for me.
I'm able to restart the service successfully when just using the host (Ubuntu 20.04) and even when the host is running Kali Linux as well, but not when it's a docker container, which doesn't make any sense to me because everything is functional with the exception of being able to start/stop the service manually.

Answer 1

I was able to get this running by cloning this repository: https://github.com/gdraheim/docker-systemctl-replacement
After cloning, I ran the following:
/root/docker-systemctl-replacement/files/docker/systemctl.py restart amazon-ssm-agent