Closed bretton closed 7 months ago
Not sure what I'm doing wrong here, qemu not liking it, perhaps the cdrom needs to be virtual device or something?
Searching sd[abcd] and nvme
Detected drives: sda
sdb
sdc
sdd
Adding SATA drives...
Adding SATA drive: sda
Adding SATA drive: sdb
Adding SATA drive: sdc
Adding SATA drive: sdd
Configured disks: -drive file=/dev/sda,format=raw -drive file=/dev/sdb,format=raw -drive file=/dev/sdc,format=raw -drive file=/dev/sdd,format=raw
INFO: /dev/kvm exists
KVM acceleration can be used
Starting qemu...
qemu-system-x86_64: -cdrom /tmp/depenguinme/9d3d69bd6e2e6411bcc36089499601718eba180b2e36e90c07f325b48a0e6c58.iso: drive with bus=1, unit=0 (index=2) exists
Terminated
it works to get mfsbsd up, and appeared to install FBSD14, but server hasn't booted. getting remote hands attached.
update: install worked fine, I had the incorrect interface name in ansible script so no network came up
Install success with correct NIC and using depenguinme-ansible script with 4 disks configured in inventory and qa.sh as source
root@server1:~ # zpool list
NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
zroot 880G 1.37G 879G - - 0% 0% 1.00x ONLINE -
root@server1:~ # zpool status
pool: zroot
state: ONLINE
config:
NAME STATE READ WRITE CKSUM
zroot ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
ada0p4 ONLINE 0 0 0
ada1p4 ONLINE 0 0 0
ada2p4 ONLINE 0 0 0
ada3p4 ONLINE 0 0 0
errors: No known data errors
I wanted a different setup with the disks, two mirrored and two mirrored, raid10 type setup. But this is fine, will try with raidz too.
it works to get mfsbsd up, and appeared to install FBSD14, but server hasn't booted. getting remote hands attached.
update: install worked fine, I had the incorrect interface name in ansible script so no network came up
Wanted to tell you this morning (you already figured it out it seems), that is is most likely fstab or em0 vs igb0 or similar issue. Great work!
Wanted to tell you this morning (you already figured it out it seems), that is is most likely fstab or em0 vs igb0 or similar issue. Great work!
Thanks. Can confirm passing in raid10
works for unattended bsdinstall process, else raidz1
.
root@server1:~ # zpool status
pool: zroot
state: ONLINE
config:
NAME STATE READ WRITE CKSUM
zroot ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
ada0p4 ONLINE 0 0 0
ada1p4 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
ada2p4 ONLINE 0 0 0
ada3p4 ONLINE 0 0 0
errors: No known data errors
Updated docs and in ansible repo to be clearer about interface names or zpool parameters. Will do a final test and then can see about merge.
Using script for recovery purposes is more practical for wider number of servers now, can reassemble existing pool, send it somewhere or other recovery options.
All looks good with latest test, ansible script too.
if someone else wants to test:
wget https://depenguin.me/qa.sh
chmod +x qa.sh
./qa.sh <url-to-pubkey>
otherwise code review
Just wanted to say: It worked for me on a Hetzner server rented through their server auction.
System:
sda
and sdd
)sdb
and sdc
)I wanted to install FreeBSD on the two SSDs as mirrors (v0.0.13 only offered me the first two, sda and sdb as ada0 and ada1). With v0.0.14 this was easily achieved with
conf_disks="ada1 ada2"
Many thanks for your work on this excellent script!
I'll take this live for now, can improve code in updates
@bretton Sorry, I didn't have the cycles to test or review.
Question: Any specific reason, why disks isn't created as a bash array but instead concatenated and converted?
Question: Any specific reason, why disks isn't created as a bash array but instead concatenated and converted?
I couldn't get add_drive_to_disks() to work, qemu gave errors and didn't see the extra items, but it worked if I made a string first then turned that into array
the if/else after calling add_drive_to_disks can also be simplified into single statement I think
The disk check logic has been tested on a few systems and works to identify nvme or sata disks.
I don't have a 4 disk instance yet, however the file is available for testing at https://depenguin.me/qa.sh
testing required before merge.
https://github.com/depenguin-me/depenguin-run/issues/60 applies