Which specific ones should it need?
# lsmod | grep qla2xxx
tcm_qla2xxx 32768 1
target_core_mod 335872 4 tcm_qla2xxx,iscsi_target_mod
qla2xxx 634880 1 tcm_qla2xxx
scsi_transport_fc 65536 3 bfa,qla2xxx,tcm_qla2xxx
Anything else?
Yes, I literally googled every single thing over the last two weeks and saw one where someone asked if that was an issue. I was just trying that to see if it was the issue.
# targetcli
targetcli shell version 2.1.fb49
Copyright 2011-2013 by Datera, Inc and others.
For help on commands, type 'help'.
/> /qla2xxx create naa.10008c7cffc7ef00
Could not create Target in configFS
/> exit
Global pref auto_save_on_exit=true
Configuration saved to /etc/target/saveconfig.json
did you try tabbing to see if maybe you type the naa id incorrectly?
Yes. I tried copy/paste and typing manually. I have 4x 2 port 16gb cards in it. I literally tried every one of them.
I tried contacting Marc at Datera who wrote the guide I originally looked at. He hasn’t answered yet. Hope he may. I’m sure they’ve come across this before.
http linux-iscsi - org/wiki/Fibre_Channel
Give fedora a try?
Before I go too far down the road, my hope is to use CLVM between 6 servers and this server with all the storage so I can do some failover across them all. Each of them should be able to create LVMs as needed on the block device. Not sure if that’s entirely possible but I hope so otherwise for this to work, all of my VMs will have to be converted from LVMs to files.
is that the literal output or did you edit out the IDs?
even if you didnt use FC HBAs… you could use iSCSI with targetcli and still have block storage for your servers.
I edited the IDs.
# targetcli
targetcli shell version 2.1.fb49
Copyright 2011-2013 by Datera, Inc and others.
For help on commands, type 'help'.
/> /qla2xxx info
Fabric module name: qla2xxx
ConfigFS path: /sys/kernel/config/target/qla2xxx
Allowed WWN types: naa
Allowed WWNs list: naa.10008c7cffc7ef00, naa.10008c7cffc7ef01, naa.10008c7cffc52800, naa.10008c7cffc52801, naa.10008c7cffc7dc00, naa.10008c7cffc7dc01, naa.10008c7cffc58b00, naa.10008c7cffc58b01
Fabric module features: acls
Corresponding kernel module: tcm_qla2xxx
/> exit
Global pref auto_save_on_exit=true
Configuration saved to /etc/target/saveconfig.json
It’s a bit late to consider anything other than FC at this point. I have about $2K into all of these cards. It was no cheap endeavor
is selinux off?
Absolutely. I kill that first thing. It seems I use too many things it disapproves of and have not been successful in making anything play nice with it.
# getenforce
Disabled
Do you have a list of all the packages you had to install on Fedora to make it work?
other than targetcli… i didnt have to install anything else, but b/c of zfs and some other “nice to haves” here is the whole install history
I’ve never used zfs, always ext4. Any major differences?
many, but thats a topic for another thread
2 Likes
I’ll google that piece.
What I mentioned with CLVM and what I want to do, see any issue with that? I have a 20tb disk I have that I wanted to make available for this purpose. I know it’ll mount as a regular device but can I setup multiple LVMs on it like that over FC?
I havent used CLVM, but I have used regular LVM with multiple proxmox nodes on shared DIY FC SAN storage. Just cant do thin lvm b/c there are multiple nodes.
i plan to look into CLVM later.
just configure the drive partition you want shared as a block device in targetcli. and share it over the FC HBAs.