UPDATE
After the miserable MOBO failure, I've reinstalled the system again, but, this time, the "install windows as usual" didn't work anymore. This is a symptom that somwthin was "strange" with first run. Anyway there is an hardcore a-la-slackware-linux solution for installation. I've updated the how-to accordingly.

premise

ok, yes, it's a microSD I know...
anyway this is the how-to for anyone in need.
I've wasted something like 2 days to figure out how to do this, despite being this a supported and certified solution. HP support has been partially useful. Infact Intelligent Provisioning (automatic provisioning for a given OS) is broken.

preparation

• remove all of your disks, if you have disks into the arrays you will not be able to install on the microsd <- yay! this is 2017 and still we play these games
• enter bios settings and be sure that USB3 is disable (you can reenable later) and VID is enabled
• make a handy hyper-v bootable usb or you will be out of luck

VID is the "virtual installation disk" and contains all the drivers required for the dl380.

installation

• enter the bios one-time boot menu end select to boot from the USB pen
• when the windows setup appears, choose the little link on the lower-left corner: "repair pc", then enter toolboxes and run a DOS command shell
• use diskpart to setup partitions on the microSD: verify which disk is your microSD, I consider it here as disk 0

diskpart
list disk 
select disk 0
clean <- this erases the disk!!!
convert gpt
create partition efi size=100
format quick fs=fat32 label="System"
assign letter="S"
create partition msr size=16
create partition primary 
format quick fs=ntfs label="Windows"
assign letter="W"
exit

• from command line:

powercfg /s 8c5e7fda-e8bf-4a96-9a85-a6e23a8c635c
dism /Apply-Image /ImageFile:D:\install.wim /Index:1 /ApplyDir:W:\ <- MIND that D: is your USB pen letter, adjust accordingly

• wait for the extraction of the entire OS on the microSD, then

bcdboot W:\Windows /s S:

• reboot, automatic, several times, just remove the USB and let the server start the normal way.

windows will boot and ask for an admin password. mind that this installation is not localized: just use letters and numbers you are able to recognize once you will adjust your localization.

final thoughts
It is damned simple straightforward, when you have discovered:

• the right combination of bios settings which correctly allows you to install
• you have understood that Intelligent Provisioning, the HP tool for setting up bios for you, is broken, will create an unbootable NTFS efi partition and will broken your USB media.

hope this will save time to someone else!

bye,
M

PS: used guides: partitioning, installation

@scottalanmiller , all

hello.

I'm an almost 40 yo guy from Italy. Had a career as machine vision consultant, then I've started to work for a SMB (50 people including warehouse) in the glass market.

It's 1 year since I've started the job. In theory the tech office of the company asked for a coder, but it emerged out that they have no IT staff, no IT at all (but a xeon ws running as a server) and I've started scratching my head in order to remove the huge chaos in the company.

Basic knowledge on linux and kvm, mostly a coder in Python and C++. Always struggling with the "IT-is-just-a-cost" and "do-it-ourselves-will-cost-less" mood in a 13.000.000 € revenue company.

Recently I've started reading spiceworks, where I've met some of the guys who also post here (notably SAM). Then started looking at some posts here too!

ok, not involved with buffalo, just a customer of theirs. but this is my story.

a couple of years ago I wasn't aware of Buffalo as a NAS vendor. I've found their appliances in the current company.
Today , back from my citrix event I've received a mail from the NAS daily check: SMART has detected "6 bad sectors" in one of the 4 3TB disks. <- they sell populated units in Italy.

Now the NAS is still covered by the default 3-year warranty, so I've called the tech support. Yes, I'm in Italy and they provide phone support localized in my language in the 9x5 timeframe. This is great, ok, but what about L1 support... everyone knows it: "hey did you check the power plug? oh, it is plugged...err.. sorry I've to ask L2 for this."

No! As with other enterprise grade companies, Buffalo L1 is really skilled people, I've asked them for other issues and always L1 solved them, being them replica issues, firmware issues etc...

OK, back in topic. I call them...
me: my daily SMART check mailed me about 6 bad sectors in 1 of the 4 disks: disk is still working and raid is not degraded but I would like to have a feedback (I actually wanted them to replace the drive).

buffalo: please give me your S/N... ok you buyed a populated NAS and it is still under warranty period. I suggest to replace the drive ASAP before it will deteriorate. Please, send back the info I will ask you by e-mail and we will send a new HDD in 24 hours. you have 15 days to return the broken one.

Honestly, I'm not aware of a vendor in the SMB/ROBO/SOHO italian market, which:

• covers every NAS with a 3+2 year warranty
• sells you populated units covering HDD damages
• has a support center close to you, in your country, givin 9x5 support to both resellers and end users
• provides 24h replacement at vendor expense (including the returned unit)
• does this even for small 2-bay units!

they HW is not so great and the SW quite old and clumsy, but this kind of support/service is great for vendor standards in my country and always lets me think about buying a new one.

well... at least when you can't afford a SAM-SD of course

LATEST NEWS:

yesterday, after a number of tests VM did freeze at reboot command rather than shutdown. As a last resort I've chosen to format SD/rest bios and restart from scratch.

My installation procedure did not work anymore... hell 2 times was right the third has failed...

called HPE this morning, they have asked me to do some tests via Intelligent Provisioning. IP. chrashed with missing linux kernel file (IP is based on linux)

they will start mobo substitution procedures tomorrow morning.

F*#@ wasted 15 days.

15 D.A.Y.S.

ever seen a failing server MOBO? I've. And I've provisioned just 3 servers in my life!

@oksana you should mention here that this is in violation of MS eula

and after a lot of episodes, I've fixed the thing.

please let me state this again: I HAVE.

I.

Not the HPE support, not the reseller tech support.

I HAVE.

I = An almost idiot ex-embedded sw developer illogically cast into the sysadmin role of a non-sense company!

TL; DR:
there was a mismatch between OS version, bios version, controllers firmware version, controller drivers version. Selecting the right combination has (apprarently) fixed the issue!

how storage controllers affect reboot signals is still a mistery... but actually all was related to the smart paging I mentioned before. Moving it from the default location caused all the issues.
I have discovered the "bug" recreating a new vm with copy-pasted hdd but with hypervisor defaults... 2 twins VM one running one not... only difference: location of the smart paging file. XD

http://www.gsmarena.com/twitter_sues_the_us_department_of_homeland_security_for_trying_to_break_the_first_amendment-blog-24425.php

It depends on what you want: virt-manager is considered the de-facto standard with few servers and VMs. Windows support is a mess in my experience. My workstation is a win10 machine with a linux VM constantly spawned, as my company uses win10 everywhere but I need a lot of linux stuff. So, when I need to go back to VM admin I simply open virt-manager in the linux VM on my workstation <- hoestly I do this rarely as my machine are created and kept running all the time.

Another tool for lightweight admin, still a bit lacking in features, is webvirt manager. It is a web app that you can even virtualize on a KVM host, then, you can connect the webvirt VM to the host (a la XOA).

I've use it for a while at the beginning but then I've gone back to virt-manager.

Moving to a more complex setup you could mind about ovirt: it is an orchestrator a la vServer and it requires a lot of stuff installed on top of libvirt-kvm. So, unless you need something like a data-center pane, avoid it. It will requires a lot of resources and setup.

just for the curious, I'm trying to configure a couple of VM with ansible and salt just to compare them.
here is my diary with the gists.

notes:

• this is a work in progress
• this is really n00b stuff: I'm a coder not a sys admin
• I'm not the kind of man than reads the full doc and then start code with order <- I rather create a lot of chaos and then I reorder and formalize after docs
• the document is a sort of diary updated live in these days whenever I've time.

bye bye

@stacksofplates yup! One of the few using golang
I've enjoyed it too once!

try this:

# initializing agi variables
declare -a array
while read -e ARG && [ "$ARG" ] ; do
 array=(` echo "$ARG" | sed -e 's/://'`)
 export ${array[0]}=${array[@]:1}
#next added line for debug only. comment this out in prod.
 echo "${array[0]}=${array[@]:1}"
done

if you want to "slice" an array use the syntax: ${array[@]:from:to}, not providing the 'to' arg means go up to the end of array.
source here.

@Dashrender said in Has anyone setup an API for an internal application:

@matteo-nunziati said in Has anyone setup an API for an internal application:

@JaredBusch so currently the file is pushed to a server placed at customers' sites?!
You can write some http API but still you need to set the server.
Can't you othetwise keep the sftp server at the source and let customer use any ftp client (even the browser) to download it? Basically this reverse the process snd customers pull the file.

I believe that Google removed FTP from the browser.
https://www.coywolf.news/productivity/chrome-removing-ftp/

Yep! still you can use file explorer for sftp if you reverse the ftp architecture. The question is: how much effort has to be put if you have to redo the app in next months?!

@JaredBusch said in Has anyone setup an API for an internal application:

@stacksofplates said in Has anyone setup an API for an internal application:

Honestly I'm not sure what you're going to get doing it that way over just normal SFTP. It would be essentially the same thing just over HTTPS instead. The advantage would come in if the remote ends had a way to consume what you were sending instead of just getting it as a file. I can't speak to EDI because I have no experience with that.

Because I do not have an FTP server. The software sends to other people's FTP servers.

But now the customer's customer does not have anything either and they want to pull the data from us.

To me, in the modern world, this means an API to connect and pull their data.

Well if you reverse the process and go for a pull it is easy to build a small REST API by anything to serve the documents and download them. And the client can be the browser or any simple script with curl if you have to automate via cron/ANY-SCHEDULER-IN-THE-OS.

Btw I remember you had .NET developers. You can build a simple auth'ed app with ASP.NET with VB.NET. It will fit your needs and it seems the easiest road, you just need to define how to store the auth credentials. and if, server side there should be a service listening for someone dropping the new CVS now and then.

@JaredBusch so currently the file is pushed to a server placed at customers' sites?!
You can write some http API but still you need to set the server.
Can't you othetwise keep the sftp server at the source and let customer use any ftp client (even the browser) to download it? Basically this reverse the process snd customers pull the file.

The issues seen in Italy:

New virus: no cure you are on your own. Too weak means you die.
Reason you die: std side effects of unthreated flu on weak people (mostly old people etc..). Covid being able to go deeper in the lungs causing primary type pneumonia (primary=directly caused by virus not by the weakening and by subsequent bacteria installation) which is dangerous even for young people.

Main issues: high diffusivity rate, high number of people to require machines to breath. Not enough machines in Italy.

Basically covid is more dangerous than flu but not so dangerous. Unfortunately the sizing of our hospital facilities is not enough to host all the people in need. Therefore, before we have to let people selectively die, government has imposed a quarantine.

@scottalanmiller we are now officially in nation-wide quarantine. You can go out only for relevant reasons.
This all mostly because our hospitals are close to their max cap.

The only service I've seen winning hand off wrt a dedicated infra is firebase. Developing apps for super small shops on firebase (the platform not the db only) is really convenient. Also there are a number of cases where developing serverless apps with pure functional back-ends is cheeeper. Butwe are talking custom software deployment not standard software montly billing. In the latter case I agree that could is really expensive for the average SMB.

@scottalanmiller said in Reconsidering ProxMox:

@Pete-S said in Reconsidering ProxMox:

@matteo-nunziati said in Reconsidering ProxMox:

Don't forget openbsd:
(https://www.openbsd.org/faq/faq16.html)

That is bhyve I believe. Vmm is the kernel module of bhyve.

Ah, that would explain why I didn't know about it

Not exactly, look at this presentation from 2016 (pdf)
https://bhyvecon.org/bhyvecon2016-Mike.pdf

@scottalanmiller said in Reconsidering ProxMox:

@jmoore said in Reconsidering ProxMox:

@black3dynamite Ok thanks. So it looks like another virtualization option. What advantages and disadvantages does it have with KVM?

The only production virtualization platforms are ESXi, Xen, KVM, Bhyve, and Hyper-V. That's it. And Bhyve might as well not exist.

All products on the market are these products. They may come in different packages, but at the end of the day, every solution is one of these.

Don't forget openbsd:
(https://www.openbsd.org/faq/faq16.html)