On an Amazon Linux (RHEL + CentOS fusion) OS server (this is actually an EC2 instance within a EMR cluster 4.5.0 AMI if that helps), we can see –
$ ls /etc/init.d/ acpid cloud-init-local instance-controller ntpd ….. $ ls /etc/init/ control-alt-delete.conf hadoop-yarn-resourcemanager.conf hive-metastore.conf …..
Further we can also see the following –
$ sudo stat /proc/1/exe File: ‘/proc/1/exe’ -> ‘/sbin/init’ $ /sbin/init --version init (upstart 0.6.5)
which tells us that the Amazon Linux based EC2 hosts are using upstart init system during boot process.
Based on this, my understanding is that –
- Different daemons or services on this host are using either upstart or systemd.
- systemd and upstart init systems coexist together and both of them are active at any given time. Is this possible? Is this understanding correct (including the verbiage)?
If yes, can you please let me know how can we achieve this (I do not see any ‘systemd’ named script on any of the EC2 hosts)?
Note : I think this might be a classic repeated question, but I am a little confused with multiple articles with inconclusive answers.
Ref –  https://forums.aws.amazon.com/message.jspa?messageID=731256 (“Systemd is not a one-to-one drop-in replacement.“)