As mentioned, the idea is to focus on those servers where only one disk is available and no external disk drive is possible to be attached. Notice that assisted installer, which is part of the ZTP flow, leverages the coreos-installer utility to write RHCOS to disk. Therefore, if we boot from a pre-installed RHCOS on the device, the utility will complain because the device is in use and cannot finish the process of writing. Then, the only way we have to run the full pre-caching process is by booting from a live ISO and using the factory-cli tool from a container image to partition and pre-cache all the artifacts required.
⚠️ RHCOS requires the disk to not be in use when is about to be written by an RHCOS image. Reinstalling onto the current boot disk is an unusual requirement, and the coreos-installer utility wasn't designed for it.
As a requirement before starting the partitioning process with the factory-cli, we need the disk to be not partitioned. If it is partitioned we can boot from an RHCOS live ISO, delete the partition and wipe the full device. See the process below for deleting a partition of the nvme0n1 disk:
[root@snonode /]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 93.8G 0 loop /run/ephemeral
loop1 7:1 0 897.3M 1 loop /sysroot
sr0 11:0 1 999M 0 rom /run/media/iso
nvme0n1 259:1 0 1.5T 0 disk
└─nvme0n1p1 259:3 0 300G 0 part
[root@snonode /]# fdisk /dev/nvme0n1
Welcome to fdisk (util-linux 2.32.1).
Changes will remain in memory only, until you decide to write them.
Be careful before using the write command.
Command (m for help): d
Selected partition 1
Partition 1 has been deleted.
Command (m for help): w
The partition table has been altered.
Calling ioctl() to re-read partition table.
Syncing disks.
[root@snonode /]# wipefs -a /dev/nvme0n1
/dev/nvme0n1: 8 bytes were erased at offset 0x00000200 (gpt): 45 46 49 20 50 41 52 54
/dev/nvme0n1: 8 bytes were erased at offset 0x1749a955e00 (gpt): 45 46 49 20 50 41 52 54
/dev/nvme0n1: 2 bytes were erased at offset 0x000001fe (PMBR): 55 aa
/dev/nvme0n1: calling ioctl to re-read partition table: Success
[root@snonode /]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 93.8G 0 loop /run/ephemeral
loop1 7:1 0 897.3M 1 loop /sysroot
sr0 11:0 1 999M 0 rom /run/media/iso
nvme0n1 259:1 0 1.5T 0 disk ❗ It is recommended to use NVME disk in your servers.
A second prerequisite is being able to pull the quay.io/openshift-kni/telco-ran-tools:latest that will be used to run the factory-cli tool. The image is publicly available in quay.io but if you are in a disconnected environment or have a corporate private registry, you will need to copy the image there so it can be downloaded to the server.
[root@snonode /]# podman pull quay.io/openshift-kni/telco-ran-tools:latest
Trying to pull quay.io/openshift-kni/telco-ran-tools:latest...
Getting image source signatures
Copying blob 97da74cc6d8f done
Copying blob 8700ab5d88a1 done
Copying blob d8190195889e done
Copying blob b250919afbec done
Copying blob 5f6dcacfa818 done
Copying blob 6b1b76a0904e done
Copying blob cf95c38e62d8 done
Copying blob 86cf6373f6c4 done
Copying config 4b99f87ac7 done
Writing manifest to image destination
Storing signatures
4b99f87ac7872f8b6cdfea50ec0aed44d5145a33b171e22c33b866ce15759c72[root@snonode /]# podman run --rm quay.io/openshift-kni/telco-ran-tools:latest -- factory-precaching-cli -v
factory-precaching-cli version 20221018.120852+main.feecf17Finally, we need to be sure that the disk is big enough to install OpenShift and also precache all the container images required: OCP release and day2 operators. Based on our experience:
- If you are going to precache only OCP release artifacts a 100GB partition is enough
- If you want to precache both OCP release artifacts and the DU profile day2 operators, around 250GB of disk space is required.
- Lastly, take into account that OCP installation requires a minimum 120GB to be installed. So, you need to add this size to precached partition.
⚠️ The factory-cli tool allows you to precache any other container image you will need for your environment. Then, make sure that the size of those extra images is taken into account in the sizing exercise.
Let's start with the partitioning, for that we will use the partition argument from the factory-cli. See here all the options we have:
[root@snonode /]# podman run --rm quay.io/openshift-kni/telco-ran-tools:latest -- factory-precaching-cli partition --help
Partitions and formats a disk
Usage:
factory-precaching-cli partition [flags]
Flags:
-d, --device string Device to be partitioned
-h, --help help for partition
-l, --label string Partition label (default "data")
-n, --num int Partition number (default 1)
-s, --size int Partition size in GB (default 100)
-v, --version version for partitionAs mentioned in the previous section, we need our disk to be empty. Please make sure it is before starting the process. This is an example of a nvme0n1 device with no partitions:
[root@snonode /]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 93.8G 0 loop /run/ephemeral
loop1 7:1 0 897.3M 1 loop /sysroot
sr0 11:0 1 999M 0 rom /run/media/iso
nvme0n1 259:1 0 1.5T 0 disk Also, it is suggested to erase filesystem, raid or partition-table signatures from the device:
[root@snonode /]# wipefs -a /dev/nvme0n1
/dev/nvme0n1: 8 bytes were erased at offset 0x00000200 (gpt): 45 46 49 20 50 41 52 54
/dev/nvme0n1: 8 bytes were erased at offset 0x1749a955e00 (gpt): 45 46 49 20 50 41 52 54
/dev/nvme0n1: 2 bytes were erased at offset 0x000001fe (PMBR): 55 aa
⚠️ The tool will fail if the disk is not empty, as it uses the partition number 1 of the device for precaching the artifacts.
Now that we have a device ready, we are going to create a single partition and a GPT partition table. This partition is going to be automatically labeled as “data” by default and created at the end of the device.
Otherwise the partition will be overridden by the coreos-installer. A custom partition label can, however, be specified using the --label option. This custom partition label will also need to be specified as arguments
to the --extract-ai.sh and --extract-ocp.sh utilities for installation, as described in the ZTP precahcing doc.
❗ The coreos-installer requires the partition to be created at the end of the device and to be labeled as "data". Both requirements are necessary to save the partition when writing the RHCOS image to disk.
Notice that the container must run as privileged since we are formatting host devices. Also, mount the /dev folder so that the process can be executed inside the container. Finally, notice that the size of the partition is 250 GB since we expect to precache also the day-2 operators required for the DU profile.
[root@snonode /]# podman run -v /dev:/dev --privileged --rm quay.io/openshift-kni/telco-ran-tools:latest -- factory-precaching-cli partition -d /dev/nvme0n1 -s 250
Partition /dev/nvme0n1p1 is being formatted
[root@snonode /]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 93.8G 0 loop /run/ephemeral
loop1 7:1 0 897.3M 1 loop /sysroot
sr0 11:0 1 999M 0 rom /run/media/iso
nvme0n1 259:1 0 1.5T 0 disk
└─nvme0n1p1 259:3 0 250G 0 part In order to verify the proper status of the partition we can query using the gdisk tool included with RHCOS. See that:
- The device now has a GPT partition table
- The partition is actually using the latest sectors of the device.
- The partition has been correctly labeled as 'data'
[root@snonode /]# gdisk -l /dev/nvme0n1
GPT fdisk (gdisk) version 1.0.3
Partition table scan:
MBR: protective
BSD: not present
APM: not present
GPT: present
Found valid GPT with protective MBR; using GPT.
Disk /dev/nvme0n1: 3125627568 sectors, 1.5 TiB
Model: Dell Express Flash PM1725b 1.6TB SFF
Sector size (logical/physical): 512/512 bytes
Disk identifier (GUID): CB5A9D44-9B3C-4174-A5C1-C64957910B61
Partition table holds up to 128 entries
Main partition table begins at sector 2 and ends at sector 33
First usable sector is 34, last usable sector is 3125627534
Partitions will be aligned on 2048-sector boundaries
Total free space is 2601338846 sectors (1.2 TiB)
Number Start (sector) End (sector) Size Code Name
1 2601338880 3125627534 250.0 GiB 8300 dataNow, that we have verified the partition was created as expected. We can mount the device into /mnt.
⚠️ It is highly suggested to mount the device into /mnt since the mounting point is taken into account during ZTP preparation stage.
Notice that the partition is also formatted as xfs.
[root@snonode /]# lsblk -f /dev/nvme0n1
NAME FSTYPE LABEL UUID MOUNTPOINT
nvme0n1
└─nvme0n1p1 xfs 1bee8ea4-d6cf-4339-b690-a76594794071 Then execute the mount in the terminal, there is no need to add the mount point in the /etc/fstab since we are running a live ISO.
[root@snonode /]# mount /dev/nvme0n1p1 /mnt/
[root@snonode /]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 93.8G 0 loop /run/ephemeral
loop1 7:1 0 897.3M 1 loop /sysroot
sr0 11:0 1 999M 0 rom /run/media/iso
nvme0n1 259:1 0 1.5T 0 disk
└─nvme0n1p1 259:2 0 250G 0 part /var/mnt❗ Notice that the mount point is actually /var/mnt. This is because the /mnt folder in RHCOS is basically a link to /var/mnt.