Mount volume permanently

List volumes

root@server:/# blkid

/dev/xvda1: LABEL="cloudimg-rootfs" UUID="ef263917-4ffc-4c36-880c-ae41d52b0d8e" TYPE="ext4"
/dev/xvdf: UUID="2c21a384-9e0e-4b44-b8d1-ceb452e8cc5c" TYPE="ext4"

root@server:/home/ubuntu# lsblk

NAME    MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda    202:0    0  32G  0 disk
└─xvda1 202:1    0  32G  0 part /
xvdf    202:80   0  32G  0 disk 

Permanently mount

Mount volume xvdf to /var/lib/mysql

root@server:/# mount /dev/xvdf /var/lib/mysql

Recheck after mount

root@server:/# lsblk

NAME    MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda    202:0    0  32G  0 disk
└─xvda1 202:1    0  32G  0 part /
xvdf    202:80   0  32G  0 disk /var/lib/mysql

Mount volume permanently - even after rebooting

root@server:/# vim /etc/fstab

UUID="2c21a384-9e0e-4b44-b8d1-ceb452e8cc5c" /data ext4 defaults  0 0

root@server:/# mount -fav
/                        : ignored
/var/lib/mysql           : already mounted

We must config fstab (permanent mount) based on UUID or LABEL like this

LABEL=cloudimg-rootfs                       /                      ext4  defaults,discard  0 0
UUID="c46cf311-d31b-41ce-bce5-5d8ad0a6b109" /var/lib/elasticsearch ext4  defaults,nofail  0 2
UUID="2c21a384-9e0e-4b44-b8d1-ceb452e8cc5c" /data                  ext4  defaults  0 0

DON'T config based on device name like this fuck

LABEL=cloudimg-rootfs	   /	                  ext4	defaults,discard	0 0
/dev/nvme1n1p1             /var/lib/elasticsearch ext4  defaults,nofail     0 2
/dev/xvdf                  /data                  ext4  defaults            0 0

Mountfuck

Entire fucking dir in fuckin /var/lib/elasticsearch

Root cause: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/nvme-ebs-volumes.html

EBS uses single-root I/O virtualization (SR-IOV) to provide volume attachments on Nitro-based instances using the NVMe specification.These devices rely on standard NVMe drivers on the operating system.These drivers typically discover attached devices by scanning the PCI bus during instance boot, and create device nodes based on the order in which the devices respond, not on how the devices are specified in the block device mapping.In Linux, NVMe device names follow the pattern /dev/nvme<x>n<y>, where <x> is the enumeration order, and, for EBS, <y> is 1.Occasionally, devices can respond to discovery in a different order in subsequent instance starts, which causes the device name to change.

So, if we use NVMe disk for some new types of AWS EC2, please note that the device name is nearly randomize after each reboot. It means if we have 2 NVMe disks on one EC2 vm, so we cannot know which device name delegate to which real disk.

/dev/nvme1n1p1
/dev/nvme0n1p1

Reference

1. https://askubuntu.com/questions/45607/how-to-mount-partition-permanently/45618#45618
2. https://docs.oracle.com/cloud/latest/computecs_common/OCSUG/GUID-075D8B13-A089-4A81-BB5C-DD7B09995C47.htm#OCSUG276