Starting drivers. Please wait, this may ... take HOURS!!!

[WOPS]

Hi

My DL380 Gen9 take hours on point "Starting drivers. Please wait, this may ..." to boot up or go on System Utilities.

BIOS Version: P89 v.2.72
iLO4: 2.62
Smart Array P440ar: 5.04, with local disks no SAN

There is no MicroSD Kit installed.

I didn't find any solution on internet, I also set the BIOS to default still problem exist.

Any ideas?

Regards, Francesco

[rmay_bk]

Any PCIe cards?  NICs?  Disk controllers?  Try pulling them out and see if the behavior changes.

[opoplawski]

I'm seeing similar, and apparently so are others: https://www.running-system.com/hpe-starting-drivers-please-wait-bios-p89-v2-40-spp-201704/

[Moderator edit: Removed the broken link.]

[Momo55]

I am having this problem for a while. 

Do we have a solution to this problem?

[Jagger_Systems]

Upgrading to ilo4_276.bin resolved this for us.

Many thanks

[nthlevel]

updating to the latest ilo didnt work for me, still takes hours to boot.

Hours starting drivers and then further more time waiting for power and thermal calibration

[BRE]

Same Problem here - update iLO to 2.76 did'n solve the problem - any further solutions would be greatly appreciated.

Regards,

Beat

[Jagger_Systems]

Have you tried powering off the blade, removing the blade from the chassis to reset it completely (i'd give it a minute from doing this), then power it back on?

[kira12]

Hello,

i have this issue with ilo 2.76 and 2.77 The P440ar has 7.00 and the BIOS is 2.76 and the Problem still exist. I try it with a minimal configuration but nothing changed. I have this Problem since a upgrade to 2x v4 CPU. replace against the v3 CPU did not solve the Problem.

Hopefully HPE has an Solution and will share with us...

best regards ré

[Bigbang001]

Hello,

i have same issue with ilo 2.78 and The P440ar has 7.00 and the BIOS is 2.80 with 1x CPU v4. I have this Problem since I installed vmware 7.0.2 in an SD card and reboot!

I even wait for 4 hours and nothing happened.

Any solution?

[kira12]

Hello Bigbag,

i solve it with using a H440 instead of the P440ar. After change the Controller an update the BIOS to 2.80 it works and it swap back to V4 CPU and the

P440ar...

best regards ré

[SomeWhoCallMeTi]

Following for updates.
I'm getting the same thing on my DL360 g9. "Starting drivers, please wait" takes about 15-30 minutes, then moves to Power and Thermal Calibration, where it sits for another 30-60 minutes depending on its mood.
I have updated iLo, and not seen any improvement. I have tried to update the P440ar, but the update fails. I don't have warranty coverage any longer, so I can't download a BIOS update. I have also run the latest NIC software updates.
I plan reboots to occur in the dead of night so nobody notices. It works, but isn't helpful for my beauty sleep.

[SomeWhoCallMeTi]

Update...

After downloading the smart array controller software update (cp032801), and the array flash update (cp039995), I have had a significant improvement in boot time. It's back to a normal 4 minutes or so. The flash update will NOT run without installing the array controller software update first. 

[JB-LH]

I, too, have a server that does this. Though, it's in production, so I can't do any troubleshooting at the moment.

However, one of the reader comments at the link posted by "opoplawski" on January 23rd, 2020, caught my eye.

https://www.running-system.com/hpe-starting-drivers-please-wait-bios-p89-v2-40-spp-201704

"Solution was to change the storage option in the Bios from the default “Scan all targets” to “Scan Configured Targets only” prior to the update."

I'd be interested in knowing if this worked for anybody.

Thanks!

[Moderator edit: The above link is no longer valid.]

[AxelThirud]

Storage options was already set to "Scan configured targets only" on one of our servers with this problem

Looking at the boot options, there was a "VMWare ESXI" entry as the first boot option. I put the SD Card there instead (where my esxi is installed) and the server booted normally. (this setting can also be set through ILO before restarting the server)

Of course, there could be some error in the bios where the "Scan configured targets only" wasn't saved properly and that this was corrected when the bios settings where saved again after changing BIOS options....

[JB-LH]

Thanks for commenting, AxelThirud. I haven't had a chance to check my server yet, since it's still in production.

[busitech]

In our case, something about a Kingston Datatraveler Flash Disk attached to a USB port triggered this issue. A firmware component on the mainboard has some serious problems…. 2-6 hour POST tests, and extremely sluggish operation of virtual machines after startup, with unexplainable lags and gaps in processing capability of the system.

[kira12]

Hi,

what Version of Bios do you use?

best regards ré

[busitech]

We’re using P89 v2.92 and iLO 2.79 presently, but have tested several prior versions as well, with the same results. We have a P840ar installed. Strangely, while the system was having issues it was recognized as a P440ar instead. In our case, the same USB device has been used throughout the lifespan of the server, and this issue has not always been present. I’m not sure if it was an electrical issue or just a change in the partitions on the device that changed the behavior. I need to do some more testing with the device. A new device was substituted and formatted which has not yet exhibited the same issue since installation.

[busitech]

During troubleshooting, HPE gave us no information which was helpful, such as whether or not other users had experienced this. I have found much more information in the community forums. HPE said that they had seen issues when NVRAM needed to be cleared, and we reset the machine to manufacturing defaults yet saw no improvement or change. Obviously that was not the issue in this case.

[kira12]

Hi,

in my case i fix it with installing a H440 instead of an P440ar. After swaping the Controller the Server works, i update the BIOS and reswap the Controller back and it will work again.

best regards ré

[gsxrman]

I have removed all components have the embeded 4GB nic, Ram,  2x CPUs and the Ilo module and still this issue remains, HPE jump in here anytime! Server is essiential a brick at this point.

[kira12]

Hello,

i tryed with different releases of Bios to bring the Server backon...

Greetings René

[MrBluesman]

I experienced the same problem here.
F9 would not work, as it is started after this hanging process, and in my case it would not boot at all, even if I waited for 10 hours.
It just kept saying "Starting devices. Please wait this may take a few moments...." (Gen10)
The G9 servers apparently say "Starting drivers", I believe.

My solution to start the server through iLO5 has been:

On the first page of the BIOS start ("Early PCI initialisation" and stuff) - press "Reset".
At the very same point on the next start - press "Cold Boot".
Then it will boot through the hangup point.

This procedure was reproducible more than 10 times on a DL380 Gen10.
I had opened a ticket with HPE before and had never got a real answer except "clear NVRAM" (which did not help as well).

And here´s my solution to get rid of the complete problem:
When the server was up and running, I updated the firmware of all components - I mean, really all.
System BIOS, iLO, Smart Array Controller, installed HBA, hard drives, network adapters - whatever component I could find and could flash remotely within the OS.
I always updated to the most recent version, even if the main version was already installed (i.e., hard drives were HPD8(A), and I flashed them to HPD8(B)).
I have been rebooting the server another 10 times after this, and did not see the issue again. This was done May 13 2023.
Firmware versions on this machine are now:
System ROM: 2.80
iLO5: 2.81
366FLR NIC: 1.3299.0
530SFP+ NIC: 7.19.14
H241 HBA: 7.20
P408i-a: 5.61(B)
EH000300JWHPL: HPD8(B)

Maybe this helps...

[MajorTom]

Hi everybody,

actually I had the same situation w/DL380 Gen9 and BIOS P89 v3.02 and iLO v2.82. Once I removed the Riser Cage, the system booted w/o problems. Then I attached card by card and recognized that a NC550SFP was issuing the problem. Exchanged it with another one, then the problem was gone, maybe this helps.

BR Tom