Justgot myself one of these servers, just read about a critical bios-update, but never had any hp-product and it's quite impossible/frustrating when you cant even download the firmware without figuring out how to register the box (still havent figured it out).
Tried to register my HPE Microserver Gen10 but failed, could not figure it out either. Does anyone have a solution to this? No matter what it says "Entitlement does not exist". Add new device only gives me a list of devices that has autoregistered.
I used to be a fan of HP products so I recommended to my boss that we should buy a bunch of Gen10 microservers with E208i-P controllers, as Edge servers for our SMV customers. I have now wasted over 60 hours troubleshooting various issues, plus going through the nightmare it apparently is to get proper firmwares and BIOS/UEFI updates.
I had no idea HP's support policies had become this draconian. Frankly, it's the worst experience I have had with IT gear in a full decade. The firmware is buggy, and downloads are locked away under various "support options". I don't even want to think about how to get updates to these in a year or two, when they're still used at our customers but don't have an "active support contract" anymore. It's crazy! There is ONE single reason to download a firmware file for a HP storage controller: to fix a problem with a HP product I already purchased. Why the hell do I need to pay for that over and over?
Thanks to everyone here for your suggestions. Downgrading (sheesh!) after swithing to Legacy mode (double-sheesh!) solved some of the problems. After sorting out the various support contract issues (triple-sheesh!!). Except off course now we are not on the most current version, which is what HP recommends. So any future problem we report to the official Support line will now be dead in the water, because the firrst recommendation is to be on current firmware.
I regret buying these units and my boss and co-workers will be riding my a** for months because it was my suggestion. I will be recommending a "never HPE" policy instead now. Sad - I know - but seriously: I don't see this stupid crap with Lenovo, SuperMicro or DELL gear.
Frankly I think the same, more than 20 years selling and recommending HP and right now I try to run away from them, I don't understand how I can't download any BIOS or driver updates if I don't have the contracted support, even a 20e Asus board can download the drivers 20 years later, why HP follows this policy??? frankly they are crazy, it's the best way to throw away all their confidence....
The power figures might seem a bit high, but the DT01ACA3 disks are real power hogs - even in standby they use more than 1W each. The old HPT controller also does no power management whatsoever. But since all disks and the controller where unused and essentially free...
About that, could you please share the zip file of the bios update?! I still can't get my head around that site how to download it, maybe is something related with my browser I don't know, but would you be so kind to share the file please?
To download the UEFI/BIOS you need an account with linked support or warranty status. It does not matter if the warranty is still valid though.
The links ask for your login, if you click on download.
This alone makes headless usage almost impossible, because you never know if the server comes up after a reboot. Also, you can never change the hardware without keyboard and video, because the damn "are you shure you wanted to open the case and plug in a new ram module?" prompt...
Also, no messages whatsoever, just the fancy logo. This caught me completely by surprise (thus my ranting here), I spent almost an hour diagnosing this before I remembered the last thing I did before rebooting was to disable SMART on 2 of my disks (which curiously won't sleep with it enabled). And behold, upon unplugging those disks the machine boots!
All in all I'd prefer an BIOS which can be made "dumb as a rock", because I'm doing the advanced stuff (RAID etc.) without it anyway *and* modern OS almost completely reinitialize the hardware even with a working BIOS...
But no, HPE tries to be smart and falls flat on it's face (especially without iLO).
Combined with the really slow boot (and thus extra punishment in case of troubles), the Gen10 is barely adequate for a semi-professional environment: At work I want a server that boots. Period. If there's something wrong with the hardware, let the OS deal with it.
Now, iLO/IPMI/iKVM would be a great way to work around the above issues, but I'm usually happy without it. Never needed it till today, even on dirt cheap hardware which runs 24/7. But with the BIOS in its current state, an iKVM, IPMI or iLO would be nice.
Edit:
Just confirmed it - enabling smart support (smartctl -s on /dev/sd?) makes the server boot again with the disks.
You can disable SMART support via the Marvell UEFI tool, but this does exactly nothing on "unconfigured" disks - after restarting the tool, settings are lost and SMART is enabled again. I presume they are saved in the RAID metadata block on the disk(s), which is not present on unconfigured disks.
Also another nitpick:
Who at HPE had the bright idea to save 0.00002ct by installing only green LEDs (which are brighter than the sun btw.) and skipping the reset switch? I mean, a red HDD led would have broken the budget? Or one of those newfangled yellow or even a blue LAN LEDs? It's really hard telling if the server is booting when all LEDs are the same color and the LAN LED blinks twice as much as the HDD LED. Having all errors/warnings in amber isn't really helpful either, but at least you can differentiate between the power and message LED by shape.
Resetting via Power switch is also quite bad for the disks, it even get logged in smart as unexpected power loss.
Didn't notice the bug on smart, because I use smart on every hard disk, and they go to sleep fine..does omv tell you anything in syslog why it didn't go to sleep? maybe is not smart (that seems rather odd if you set it the right way) are you using plex by any chance, that would keep your hard disk always on no matter what if you are using it or not.
Completely agree with all your other picks, I feel the same especially with the leds, the memory and boot slowness not so much, unless you need to be uptime and less downtime possibile for critical usage like webserver or such (but I guess this is not the server to use for such kind of stuff).
I mean yes it's annoying, but how many times do you really change ram?
When you settled with how much ram you want, it's done, one time at configuration and you will never touch it again, if doesn't get fucked up, but again...after the hassle to go there open up the pc and change it, connect a keyboard\video to it and say yes is not that troublesome (is it stupid?! fuck yes!).
Same goes for the boot speed...I don't care too much it's running 24\7 under ups, so I hope it will stay online forever :D, if not wait once a month to reboot is not big deal (again is annoying?! fuck yes, especially because I don't understand what the heck is doing under the hood to take all that time :D, for sure makes you want to throw your ssd out of the window not being so useful :D)
The led choice is beyond me...I guess is not fashion with different leds :D or it doesn't match the new hpe logo\color scheme!?!?!? :D you know fashion style notoriously important for a server that probably will be put inside a dusty cabinet :D
Looking at the bright side at least it got a dust filter...I've got an old gen6 that it doesn't even have one with two fan blowing inside as hell...for a tower pc that probably would have eaten up more dust than a junkie, that wasn't a smart decision..but again...hp I guess decision department sometime is composed by drunk monkey :D or just they think like "this seems useful..well fuck that! let's just screw it up!" :D
@qwertz123
Didn't notice the bug on smart, because I use smart on every hard disk, and they go to sleep fine..does omv tell you anything in syslog why it didn't go to sleep? maybe is not smart (that seems rather odd if you set it the right way) are you using plex by any chance, that would keep your hard disk always on no matter what if you are using it or not.
I wasted a lot of time on this before I found out:
OMV enables smart "offline" data collection (smartctl -o on) which is generally a good Idea. Unfortunately, it's an automated scan of some sort which should run every 4 hours (the disk firmware manages this). Contrary to all other disks I know the Toshiba DTA01ATAx00 Series seems:
I hope this was only an initial scan (disks where in storage a long time), otherwise I'll just patch the setting out of /usr/share/openmediavault/mkconf/smartmontools. The power consumption logging will show what's happening...
Not often, I know. I was basically just whining at this point, working my frustration out
But IMHO its important to document these caveats and I'll have to remember them - because those points are different from basically any computer I ever owned.
I agree: With any linux OS it falls under "nice to have", for my use having iLo would be worse than not having it: The BMC which is running all the time usually consumes 5-10W power, which is expensive in the long run (because power is expensive here).
Hello, I'm new to this forum, but I have been used OMV now for a while. I received my HPE Microserver Gen10 about two weeks ago, it's running with the recent OMV release (3.0.91).
It's a nice toy so far, doing everything I want. However, this morning the server was down. It turned out, that it might be a temperature problem, The boot screen told me something like this. My server is located in the roof top of our house, which is kind of isolated, but at recent temperatures (the outside value dropped below 0C tonight) it get's cold there. Is there a workaround for this? I prefer to have it located there, because no one gets annoyed by the server's noise and there's no dirt in the room.
Thanks in advance
Christof
3a8082e126