View Issue Details

IDProjectCategoryView StatusLast Update
0000466AlmaLinux-9systemdpublic2024-05-27 11:27
Reporterjouvin Assigned To 
Status newResolutionopen 
Summary0000466: Failure to install 9.4 beta on a server equiped with a LSI SAS 1068E controller

I'm trying to install AL9 on an old Dell C6100 server which has a LSI MPT SAS 1068E disk controller. I used AL9.4 beta (it is a test system) to benefit from the reintroduced HW support for some controllers removed originally from EL9 (and sometimes from EL8). Unfortunately it doesn't work: I see an error during the boot saying that `udev` device initialization has not been completed and the service `systemd-udev-settle.service` is failed.

I tried to disable `apci` as suggested by some posts but without success. I was wondering if there is something I missed or if it is a known issue. This machine is not UEFI-capable and thus is using Legacy BIOS, not sure if it can have an impact. And the server has 2 disks configured as a mirror at the controller level.

I checked the vendor/device ID for the controller:

02:00.0 0100: 1000:0058 (rev 08)
        Subsystem: 1170:1000
        Flags: bus master, fast devsel, latency 0, IRQ 37
        I/O ports at d000 [size=256]
        Memory at fbbec000 (64-bit, non-prefetchable) [size=16K]
        Memory at fbbf0000 (64-bit, non-prefetchable) [size=64K]
        Expansion ROM at fbc00000 [disabled] [size=2M]
        Capabilities: [50] Power Management version 2
        Capabilities: [68] Express Endpoint, MSI 00
        Capabilities: [98] MSI: Enable- Count=1/1 Maskable- 64bit+
        Capabilities: [b0] MSI-X: Enable- Count=1 Masked-
        Capabilities: [100] Advanced Error Reporting
        Kernel driver in use: mptsas
        Kernel modules: mptsas

which seems to match one of the `modules.alias` entry:

alias pci:v00001000d00000058sv*sd*bc*sc*i* mptsas

OTOH, I have successfully 9.4 beta to install a Dell C6220 which has a LSI MPT SAS 2008 and is able of using UEFI.

Thanks in advance for any help.
Steps To ReproducePXE installation of AlmaLinux 9.4 beta
Additional InformationThe installation is taking ages but progressing... Yesterday I was lucky to have an installation done after 24 hours or so... (more than 15h). Very quicly, at the end of the pre installation script (in our case it used to wipe and partition the installation disk, issuing several `udevadm settle` commands that all time out), there are 64 `systemd-udev` processes that remain here forever. The exact message when `udevadm settle` times out is:

Timed out for waiting the udev queue being empty.

On the system that completed its installation in 0000021:0000024 hours, I was able to reboot it but the shutdown was delayed several minutes because of 2 `systemd-udev` processes hanged and killed after 245s.

Status of `systemd-udev-settle` during installation!

[anaconda root@os-77026 ~]# systemctl status systemd-udev-settle.service
× systemd-udev-settle.service - Wait for udev To Complete Device Initialization
     Loaded: loaded (/usr/lib/systemd/system/systemd-udev-settle.service; static)
     Active: failed (Result: exit-code) since Mon 2024-04-22 11:09:53 UTC; 58s ago
   Duration: 18.109s
       Docs: man:systemd-udev-settle.service(8)
    Process: 2566 ExecStart=udevadm settle (code=exited, status=1/FAILURE)
   Main PID: 2566 (code=exited, status=1/FAILURE)
        CPU: 10ms

Apr 22 11:07:52 systemd[1]: Starting Wait for udev To Complete Device Initialization...
Apr 22 11:07:52 udevadm[2566]: systemd-udev-settle.service is deprecated. Please fix multipathd.service,, anaconda-pre.service, anaco>
Apr 22 11:09:53 udevadm[2566]: Timed out for waiting the udev queue being empty.
Apr 22 11:09:53 systemd[1]: systemd-udev-settle.service: Main process exited, code=exited, status=1/FAILURE
Apr 22 11:09:53 systemd[1]: systemd-udev-settle.service: Failed with result 'exit-code'.
Apr 22 11:09:53 systemd[1]: Failed to start Wait for udev To Complete Device Initialization.

`dmesg | grep mpt` returns:

[ 0.016429] Device empty
[ 0.129411] Dynamic Preempt: voluntary
[ 0.129552] rcu: Preemptible hierarchical RCU implementation.
[ 0.187893] MDS: Vulnerable: Clear CPU buffers attempted, no microcode
[ 8.420796] Warning: Unmaintained driver is detected: mptbase
[ 8.428642] Warning: Unmaintained driver is detected: mptscsih
[ 8.453856] Warning: Unmaintained driver is detected: mptsas
[ 8.468331] mptbase: ioc0: Initiating bringup
[ 24.609644] mptsas: ioc0: attaching ssp device: fw_channel 0, fw_id 8, phy 0, sas_addr 0x5000c50028bb01dd
[ 24.620636] mptsas: ioc0: attaching ssp device: fw_channel 0, fw_id 1, phy 1, sas_addr 0x5000c50028baf4f1
[ 24.631986] mptsas: ioc0: attaching raid volume, channel 1, id 0

`lsblk` shows the disk device attached to the LSI controller.
TagsNo tags attached.



2024-05-05 09:11

reporter   ~0001029


I have not seen any answer to this issue opened 2 weeks ago... Any chance that this problem will be corrected during the beta or at least in the final release?

Best regards,



2024-05-27 11:27

reporter   ~0001046


I gave a try to 9.4 release and the problem remains the same (with the same logs as far as I can tell). Please let me know if I can provide any additional information that could be useful?

Best regards,


Issue History

Date Modified Username Field Change
2024-04-24 17:04 jouvin New Issue
2024-05-05 09:11 jouvin Note Added: 0001029
2024-05-27 11:27 jouvin Note Added: 0001046