OK, so I installed an Adata NVME PCIe SSD, specifically XPG SX8200 Pro 512GB 3D NAND NVMe. I installed LEAP 15.1 on it with root and home and efi-boot and swap all on the drive with around 424GB left over which I intend to use for storing my VirtualBox virtual machines. I did not create that partition or fomat it until today. I’ve been booting and using the system for the past week or so.
Everything has been fine until I created the additional partition for the remaining portion of the drive which is intended for the virtual machines. I installed a logical partition inside the extended partition (which also holds my home partition) using Gparted from the Gparted live CD. The partition is formatted as ext4. The file system label is Data6.
lsblk shows:
lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 3.7T 0 disk
└─sda1 8:1 0 3.7T 0 part /Data2
sdb 8:16 0 3.7T 0 disk
└─sdb1 8:17 0 3.7T 0 part /Data5
sdc 8:32 0 3.7T 0 disk
├─sdc4 8:36 0 10G 0 part /OldHome
├─sdc5 8:37 0 1.8T 0 part /Data1
└─sdc6 8:38 0 1.8T 0 part /Data4
sdd 8:48 0 1.8T 0 disk
└─sdd1 8:49 0 1.8T 0 part /Data3
sr0 11:0 1 1024M 0 rom
nvme0n1 259:0 0 477G 0 disk
├─nvme0n1p1 259:1 0 1K 0 part
├─nvme0n1p2 259:2 0 500M 0 part /boot/efi
├─nvme0n1p3 259:3 0 4G 0 part [SWAP]
├─nvme0n1p4 259:4 0 35G 0 part /
├─nvme0n1p5 259:5 0 15G 0 part /home
└─nvme0n1p6 259:6 0 409.2G 0 part /Data6
I created a mount point for it under / called Data6. I edited the /etc/fstab to put the new partition in with the exact same options as all my hard drive partitions, i.e., as you see here:
rhack@linux-h2ol:~> cat /etc/fstab
UUID=9c8dce57-1c43-4436-86dd-06b70db462e2 swap swap defaults 0 0
UUID=7df10d3e-99cf-4d67-8c19-b319d26e9d96 / ext4 acl,user_xattr 0 1
UUID=c3c2b83e-5d88-4da0-8abb-d97d61ea3333 /home ext4 data=ordered,acl,user_xattr 0 2
UUID=86f5a543-3e63-4401-ae20-ea8264188377 /Data1 ext4 user,exec,acl,user_xattr 0 2
UUID=b610d0cd-9066-44db-8629-4a2369d11aa5 /Data2 ext4 user,exec,acl,user_xattr 0 2
UUID=c36a110c-1a13-46e9-b275-4c5203dab88b /Data3 ext4 user,exec,acl,user_xattr 0 2
UUID=85e6b16d-c00e-4f6d-a9c1-0c61de664e50 /Data4 ext4 user,exec,acl,user_xattr 0 2
UUID=bba2cf27-d77e-4ef8-9f02-01c29b7c140f /Data5 ext4 user,exec,acl,user_xattr 0 2
UUID=7d11e52d-e270-4950-a7fe-9895bad7d49a /Data6 ext4 user,exec,acl,user_xattr 0 2
UUID=ee67b744-61c6-4892-aad4-5146d9c9e33f /OldHome ext4 user,exec,acl,user_xattr 0 2
UUID=0F45-A46C /boot/efi vfat defaults 0 0
When I boot, the system hangs for several minutes, then drops to emergency mode, and I have to log in as root and examine the journal which shows that the systemd.fsck.service start failed due to a timeout on that partition.
I’ve spent the last several hours researching the issue. I find that according to the documentation on the systemd fsck service, it is supposed to obey the file check flag in the /etc/fstab file. So I set that to 0 to skip the check to at least see if the system would boot if it skips the check. That had no effect. Systemd continues to run the fsck start and fails on that partition. The only way to get the system to complete the boot is to remove the partition from the /etc/fstab table. Examining the partition with Gparted shows nothing evidently wrong. Running a manual fsck on that partition when unmounted shows it as clean:
fsck /dev/nvme0n1p6
fsck from util-linux 2.33.1
e2fsck 1.43.8 (1-Jan-2018)
Data6: clean, 11/26820608 files, 1963437/107269632 blocks
I can manually mount and umount the partition on the file system directory with no problem as a normal user. I can then create a file on it with no problem. The problem appears to be strictly the systemd.fsck-service times on out when accessing that partition and that partition only.
I’m at my wit’s end. The primary two reasons for buying and installing an NVME SSD was to speed up boot and program loading - and to run virtual machines at better speed than on a hard drive. Of course, I can manually mount the partition as I indicated above but that’s going to be an irritation. What really gripes me is that the systemd fsck service is not adhering to the documentation where it says it is supposed to acknowledge the file check flag in /etc/fstab - and it absolutely does not. I’ve also read elsewhere that most of the workarounds for this sort of thing which are intended to stop the systemd fsck service from doing this sort don’t work either. In other words, there’s no apparent way to turn this nonsense off.
Thanks in advance for any help.