Leaderboard

as long as it doesn't block airflow in the case you will be fine. I've attached SSD's with double sided tape to inside and lived

It does bring many things in to question, I have gone cheap on some things, others not so much. my DVD/BR Rips that make up probably 90-95% of the data usage I do not want to have to replace but I can consequently I have a RAID 6 array, the critical data is on RAID1 as for losing a RAID on rebuild, its an old fear mongering article but it gets the point across is this one below. it explains the math behind what happens https://www.zdnet.com/article/why-raid-5-stops-working-in-2009/ Obviously it did not/does not stop working but the efficacy of it as data protection becomes a greater and greater issue as storage devices increase in size

Bump this topic:) So I upgrade my microserver 8 to xeon 1270v2, and decided to install active cooling I try to won on ebay Scythe Kozuti, but oops. And then read this topic and bought Noctua NH-L9i. But things about to install it on cable tire not good. So I found a guy with milling machine, and he made a "short" legs for L9i Original nocuta upside, milled - below Cooler perfectly fit on board using Noutua screws with additional spacers from tray side (hello HPE with non-standart 1155 dimensions). And don't forgot remove nuts for original heatsink from motherboard tray. Cooler connected to case fan connector with Y-splitter.

That is what I would buy if I could still find them in the UK. I did find an web article somewhere that says that the Pro models have MLC cells that have a longer lifetime than the TLC used in the EVO models and I think it also said that the higher grade models have a small proportion of the even more reliable SLC cells (used in top end business models) as a sort of cache and that the 850 pro had the greatest proportion. Another web page also helps... https://www.datenreise.de/en/ssd-comparison-differences-between-samsung-pro-and-evo-series/

Always, get everything in writing, get a follow up email and do not action unless responded to. Cover your arse, no one else will

All good. MSFT have progressively improved the re-installation process for Windows 10 so that one no longer has to enter 25-character product codes ... or retrieve disks ... and from May 2020 rely on a manufacturer's recovery partition: the installation files can be downloaded from the net by simply selecting an option. The OS recognises the hardware and its associated license and will permit limited changes e.g. new OS disk. Better still upgrading from Widows 7, 8, 8.1 to Windows 10 is still allowed FOC and will result in a permanently activated W10 license. Great for giving a new lease of life to older devices, as you have seen. In most cases upgrading an old device with an SSD for say £35 and free W10 is a sweet deal 😀 One can also (optionally) associate a Microsoft email account (MSA) with the machine license. https://www.ghacks.net/2016/08/15/link-microsoft-account-windows-10-license/ My guess is this would put you in good stead with MSFT for an extra change, say the motherboard failed.

wrong is a funny word in this context. is the drive accessible to the cables and not going to shift around? you should be good. sometimes I have been known to have a ssd just 'sitting in there' and I am not afraid because I don't transport my 20 pound pcs.

Yeah, thanks America for that one...

https://www.hpe.com/us/en/resources/integrated-systems/ilo-advanced-trial.html

As this server supports PCIe bifurcation, I think I would try one of these for dual NVMe: https://www.supermicro.com/en/products/accessories/addon/AOC-SLG3-2M2.php

Decided to buy one. Also saw ILO advanced keys are available free from HP for 2020. Got this card. Couldn’t find one that allowed 2 x M.2 NVMe SSD drives though: SilverStone Technology M.2 PCIE Adapter for SATA or PCIE NVMe SSD with Advanced Thermal Solution (ECM22) https://www.amazon.com/dp/B075ZNWS9Y/ref=cm_sw_r_cp_api_i_cqj0Eb8MDNN8S

Interesting information. Be sure to watch the video: https://www.servethehome.com/wd-red-smr-vs-cmr-tested-avoid-red-smr/

the couple of times I've been involved in this sort of thing , I was told not to delete anything until I was given the go ahead by our legal department I dont think this is being handled properly

IANAL ... and without details of the litigation but ... ... I would check that the backups are working and that all records are kept. If your management team is honest then they might end up thanking you for proving their case. If not ... and they fire you ... then I'd expect you to have an easy claim for unfair dismissal and you might well be better out of such a company anyways. Easy for me to say of course!

Good guide here: https://www.windowscentral.com/how-create-windows-10-usb-bootable-media-uefi-support

Fire up the MCT and it'll download Win10 to the USB stick for you, then boot from it and install..

Without having a detailed diagram of the house circuit and drilling it down to the specific broken appliances/devices and where they are in the circuit, we can only speculate about what happened. My money is on overvoltage (not a spike). If it were a spike, the whole-house surge suppression would've tripped. APC has a nice write-up covering all types of power problems here: https://download.schneider-electric.com/files?p_Doc_Ref=SPD_VAVR-5WKLPK_EN You can check with your neighbors if they encountered broken appliances and/or tripped breakers. If they did, the problem is likely to have come from the utility and not from within your home. Does your surviving UPS keep logs of what happened that might give you a clue? My UPS keeps a log with a timestamp of the reason why it switched to batteries (ie. self-test, low input voltage, transient events, blackouts, etc.)

Oh, so stupid! The card is not aligned! I should get a low profile bracket. It works now. Thx you, schoondoggy!

I hate to ask, but it looks like the card is not aligned properly, have you tried aligning it to the back panel and seating it?

to sorta close my end on this - after reading these comments and what I've learned elsewhere. From a NAS standpoint, I think RAID 10 is the way to go, at least for those of us with a QNAP Something like Synology's SHR2 or a unraid type solution may also be suitable. QNAP does not offer this feature. Basically I dont want to risk losing the entire array at once if I lose a single drive. Especially as my cohort of drives ages together. When I ran a W10 server using drivepool, I basically swapped a new drive into the pool every year or so. Will consider going back to a Windows Server when I do replace my QNAP at some point . so that I can go back to drivepool. As I have been reminded - Raid is not backup My serious stuff, I am keeping on onedrive as well as the NAS. Other than that , Will continue to do a routine server backup.

You could also for simplicity remove only 1 drive and turn on the computer, if it says no boot drive then you no you have the right one.

Motherboard manual? You could just take both drives out for now, install the SSD and then once you're all back up and running put the other drives back in, reformat them and use them for storage, regardless of what connector they are.

I use an external USB (Sandisk Ultra Fit) for my bootable ESXi device. Personally, I prefer an external USB than internal, since it's much easier to maintain. If the USB goes bad, or I want to test a different version of ESXi, I can easily put another USB without having to disconnect cables and opening up the server. My USB stick barely sticks out of the port, so it's very secure. Besides, it's at home, so noone will just come by and pull out the USB stick. Just make sure you take a regular ESXi configuration backup or better, script it to do the backup daily. Some folks like cloning their USB stick to another USB stick, but I find that too inconvenient. As long as I have the ESXi configuration backup, removing from the (rare) USB stick going bad should be straightforward. Note that once ESXi has booted up, it resides in memory, so the performance of the boot up device doesn't matter. You can't just update the power supply. The overall box really isn't handled for more than 180W in terms of cooling, etc. I got my server, but my iLO Enablement Kit order won't come until mid-June. This is the first HPE server that I've had where I actually had to connect a keyboard and monitor to. Very inconvenient, and I really miss iLO. The HPE SPP should only upgrade software that needs upgrading. From my experience, SPP only updates software that needs updating. I ran the March SPP, and it only found 1 thing to upgrade (the iLO firmware, which doesn't help me at this time since I don't yet have the iLO Enablement kit).

Someone may have mentioned this previously (I'm so bad at forums), but apparently Drobo has taken the WD Red SMRs off their compatible list. And there's a known bug now where the Drobo sees them as SSDs...so that can't be helping anything if it actually attempts to manage them differently. When I removed the SMRs and replaced with an older Red, it all worked again. 🤷‍♂️

B120i has support for hot plug, who says it doesn't, look at https://support.hpe.com/hpesc/public/docDisplay?docLocale=en_US&docId=c03761790 HP Dynamic Smart Array B120i Controller Specification Description Dimensions Embedded Disk drive and enclosure interface - Transfer rate Up to 6 Gb/s SATA (Serial ATA) Connectors 6 internal SATA connectors Server interface x4 PCIe 2.0 SATA speed 6 Gb/s SATA links Cache memory Zero Memory RAID has no cache memory Logical drives supported Up to 8 logical volumes (6 Physical drives) Host memory addressing 64-Bit, supporting greater than 4GB server memory space Hot plug support Yes RAID support RAID 0 (Striping), RAID 1 (Mirroring) RAID 1+0 (Striping and mirroring) But how to get it to work with hot plug or hot swap, that is another question, anyone can shed light?

My primary Gen 8 has been running since september 2014 24/7 The only thing i have noticed is the blue led is 1/4 of the brightness of my backup gen8 that one is off 99% of the time and there is a touch of browning on the clear plastic on the 24 pin atx cable on the 12v lines due to the 12v lines running load through them and warming up slightly Failure risks will be 1) PSU 2) 4GB Nand Chip that ilo uses

No. I have a RedHat 8 installed on it running from the same SSD. I'm using the OS for my apps and KVM is just for lab servers.

My setup: microserver gen8 with an Xeon(R) CPU E31260L @ 2.40GHz 16GB RAM Smart Array P420 2 disk SSD 240GB disks raid1 - OS + apps 4 3TB disks RAID6 - DATA 1 3TB external disk - torrents. I'm running plain livbirt(KVM) with kimchi interface (https://github.com/kimchi-project/kimchi) running 5 VM. Fan Speed : ( = 11)

An LSI based HBA with two external SFF-8088 connectors would allow you to connect 8 drives external. Sans Digital: https://www.sansdigital.com/tr4x12g.html SilverStone: https://www.silverstonetek.com/product.php?pid=718&area=en Build your own:

I used to use the internal MicroSD as a boot drive for ESXi. However, one day, it went bad, and it was a hassle having to disconnect all the cables and open up the server. So I decided to use an external USB stick (one of the Sandisk Ultra Fit 16GB USBs that don't stick out much). Maintaining my server was easier with an external USB stick. It also made it easier to test booting up different or new OSs. My server is in the basement where noone goes so it's not a risk having an external USB. I have a P222 controller so use RAID 1 with 2 SSDs and RAID 1 with two hard disks. I use internal SSD for ESXi datastore. It's been a number of years since I setup my server, and I haven't actually had much need for the hard disks (it turns out that the VMs I put in my server didn't need much capacity for data). If I had known I wasn't going to use the hard disks, I may not have bothered with the P222/RAID 1. A good backup may have sufficed (assuming you have the luxury of downtime for restore). A couple of other ESXi servers I built later just use NVMe (and not using RAID).

No problems with fan noise on my server with P222.

Thanks for this, also wanted to contribute that this same topology change allowed me to upgrade the firmware on my 10-port IDsonix USB 3.0 hub from 9081 to 9101. I also got the failures mentioned in other posts until I changed these port indices of the lower-level hubs. So anyone with an IDsonix 10-port, VL812B2 upgrade to 9101 is possible.

Hey, I can comment a bit on some of your question as I just migrated my homelab from esxi 6 to proxmox-ve 6.1-11 Quick intro on my setup: microserver gen8 with an Itnel Xeon E3-1230 V2 @3.60GHz 16GB RAM 1 disk 2TB WD Red NAS HardDrive on bay 1 1 SSD 240GB Crucial on bay 2 Was tired with ESXI as limitation on licence and the need to have the vcenter installed on top of my esxi to manage the vms taking much of the available RAM and also licence bound which was a nightmare whenever you need to upgrade. Long story short, I decided to move on and get rid of the esxi and install Proxmox. Was also considering https://xcp-ng.org/ but with the uncertainty of what is going on with citrix trying to kill the open source project, at least that is my feeling while reading here and there, I finally decided to go with Proxmox which, btw, is debian based which I am comfortable with and has its prox and cons that on the overall suits my use case better for now. Well, I am not an expert, just want to keep it simple and have an working lab to play with to test new techs like deploying a kubernetes cluster and playing with it and integrating with my git ci/cd etc. Back to your question. questions 1 and 2 I think that any combination would do the work. In my case as I only have 2 disks, I installed everything hypervisor on the 2TB disk on the bay1 and left the SSD for whenever I will need it to spin a VM that needs to be snappier. I also decided to deploy ZFS as RAID0 on my 2TB disk as my intent here is to get used with this FS as I am more familiar with the good old LVM and having tho create the PVS, VGS and LVS and ultimately your EXT4 partition etc. I must confess that I am enjoying the easiness of use of ZFS and how we can create Datasets with quotas to present as mount points to your application and also the sharenfs that is present on ZFS really cool stuff. Keep in mind thought that the caveat is ZFS is memory RAM hunger intensive and it'll certainly take a toll on the total number of VM's you can spin at a time which addresses you question number 3. I have 3 VM's running K8S cluster and with the ZFS it is already taking 79% of Total RAM available so the way it's going I only have room for a few more VM's to run. But as I am testing a K8S cluster here I can spin as many pods as I need which is my main goal for now on this setup. Question 5 , my fan is quiet as ever and it is not increasing the speed and it's been quieter than ever so far. Well hope I addressed some of your concerns. You may want to have a look on the xcp-ng as it seems to be a solid free hypervisor as well.

I have been looking off and on for a file sync software to sync the files between two WHS (2011 and 2012). I stumbled across Allways Sync but the reviews have been luke warm at best. I am looking for a file sync software that will run on WHS 2011 and WS2012E and sync files automatically between both severs. They are both on the same network. Thanks for any suggests.

I ordered the 1265l v2 finally on ebay and it works perfectly. I am not seeing any heat issues after running it for a few days. Hree are some readings from iLO interface. The only thing is i don't know what to do with the Celeron processor :)

The limitation is with Intel. This family of CPU can support 32GB of memory, but it is limited to 8GB per DIMM slot. I think people have tried 16GB UDIMMs, but they have not worked.

This is my first post here but I've been lurking in this forum for quite sometime now, and from what I understand, the only reason that we couldn't achieve 32gb ram on the Microserver Gen8 was because there was no 16gb unbuffered ecc ram available before. I was searching for a memory module meeting this criteria and I found this one. My question is, letting aside this ridiculous price, is there any compatibility issue that could make this ram won't work on the Microserver Gen8? Most of the threads here about memory compatibility are quite old and I didn't want to revive an old thread just to ask this, so I'm creating this one. Thanks in advance.

just use google and add site:homeservershow.com/forums/ at the end

About 1.5 years ago, I purchased this Lycom DT-120 card: https://www.amazon.com/gp/product/B00MYCQP38/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 I put a 1TB Samsung 970 EVO NVMe on it. I'm using it in my HPE ML30 Gen9, which runs ESXi v6.5 U3. Note: I boot ESXi off a USB stick and use the NVMe as a datastore, but IIRC, it's possible to boot off the NVMe with the card as well. I learned about this card from https://tinkertry.com/how-to-install-a-2nd-samsung-950-pro-m2-nvme-on-superserver .

I noticed this recently on some of my machines as well. They routinely update, and are set to download damn near everything from Windows Update, including "other Microsoft products" as well as drivers, etc. It would appear the latest feature pack or cumulative update included some updates to the Microsoft accounts. I also converted a local account to a Microsoft one and was greeted with this. I definitely like the settings sync across devices. And with an O365 Home subscription, you can't beat 1TB per user (up to 5 users).

I decided it be nice to be able to take CPU utilization away from the CPU by getting a GPU. I wound up getting the Nvidia Quadro P2200 and installed it in my HPE ML30 Gen9. I setup passthrough on an Ubuntu 18.04 ESXi VM. Plex works well with it, doing both hardware encode and decode.

Are you apologetic because you are truly sorry or that you got caught? - Said every Mom on the planet.

Yeah lets be fair, you say to their credit, if they hadn't been caught out then we'd all still be in the dark, they didn't do this through choice, someone somewhere at WD made the call not to inform the customer.

It's not a server unless it sounds like a wind tunnel

@AlexM : Sometimes RTFM is the only and last solution. Great it worked out for you. Happy patching it to build 11675023.