zENjA

For power testing -> build youself a 3 Phase Braker box with like...2 Outlets per phase... and monitor it with a Shelly 3EM.
I've used two of these in my Datacenter at work to get monitoring data for a A&B Line for 80A 3p in a fixed install.
If you have something else... I could create a quick wireing diagram, so someone could cable a box for that.
So maybe this could be a qick idea to test efficiany between 120v & 240v - or 400v if your power provider is nice to you.
Maybe I just have a verry german view on power infra... 😅
 
Yes... you have gotten a psu tester for that... I know, but this would be a small box, that you could just throw on the table and go testing.

Wonder if Floatplane can do HDR?
 
*hears Luke's cries of anguish

"15TB... more that your total household..."
Datahorder... Hold my Ingest Drive...

My recommendation: Buy a kill-a-watt (https://www.amazon.com/P3-P4400-Electricity-Usage-Monitor/dp/B00009MDBU) (or maybe the newer version?) and plug each device you plan to use into it. If you KNOW that you won't exceed a 10 amp draw then you can just connect the power strip directly into the kill-a-watt. Then, run things as you'd expect them to be during a power failure. IE: if you plan to play a game on one PC but the other is browsing reddit, do that while playing music.
 
If you can't plug everything in all at once (and maybe consider not doing that because the device WILL explode if you exceed 10A draw), just add the various numbers together. IE: PC1 + modem + router = 450W, PC2 + printer + speakers + landline = 150W means you need 150 + 450W == 600W.
 
Then go to the kill-a-watt and see what the "Watts" mode displays. That's how much power draw you have. "Right...but, how big of a UPS do I need?" How long do you want to keep your devices running in the case of a power failure? Do you plan to be home when a failure occurs, or out of the house? If you want to have everything shut down cleanly even when you are out of the house you'll need to spend a bit more and get what is called a 'smart UPS'. The 'Smart' part is basically a USB cable and some software you install that automatically shuts your PC down when the battery level gets too low. Preventing (barring bad setups/scenarios) the battery from running out of power before your devices can cleanly power off.
 
If you plan to be at home during any power outage and only need 5 minutes of power, then you can get a much smaller UPS. Let's start with an example UPS: https://www.apc.com/shop/us/en/products/APC-Back-UPS-650/P-BE650G1
 
It says it has 390W or 650VA. Well, the VA rating has to do with the internals of the unit. The closer the VA and Watt rating are the better, but the watt rating is the one you care about. So that UPS can only output 390W at max, so it's too small for your 600W load. But you could buy two of them and split up the power load between the two of them.

"Cool, but how long will it run for?" APC has a nice calculator on their website, but let's work it out ourselves. The battery is rated for 79 Watt-hours (Battery Volt-Amp-Hour Capacity). You want 600W for 5 minutes, so how many watt-hours is that? 5 (minutes) / 60 (minutes per hour) = 0.0833 (% of an hour you need) * 600W (a 600 Watt-hour battery would run your 600W load for 1 hour, assuming 100% efficiency, but more on efficiency later on) = 50 Watt-hours to run your 600W load for 5 minutes. "But wait, the website says a 390W load would only last 3 minutes?" Efficiency. I'll get to that.
 
"Oh, so that UPS *can* run my load!" No. It cannot. Think about it this way: You have a car with a top speed of 30 MPH but a 100 gallon fuel tank. It can only ever go 30 MPH (deliver 390W) but it can keep doing that for over a longer time than you need because of the 100 gallon gas tank (the 79 Watt-hour rating). Same concept applies to batteries; capacity and maximum delivery rate are two VERY different things!
 
Efficiency: Remember that VA rating? A VA is a volt-amp. A Watt is a volt-amp-second. IE: 1 volt, at 1 amp, for 1 second is 1 watt. So that UPS has a 390 Watt rating, but it built to the 650VA spec. What that actually means is that the unit *draws* 650W from the battery, but can only deliver 390W of that power to your devices due to efficiency losses in the power conversion. 390/650 = 60% conversion efficiency. So that 79 Watt-hour rating is actually: 79 * .6 = 47.4 Watt-hours. To deliver 390W for 3 minutes only needs 19.5 Watt-hours, so I'm not sure where the extra 27.9 Watt-hours go. Maybe I missed something / fail to understand something. Anyone got some insight here?
 
One other feature you MUST have: Sine-wave output. The PSUs in your computer will randomly shut down if you use a cheap/low-grade UPS that uses simulated sine-wave power output. This is because the PSU itself is doing a thing called Power Factor Correction, and unless it gets a natural sine wave on the power input, it will shut down to protect itself. Name brand UPS (APC, CyberPower, etc) will generally come with this feature as standard. Be sure to ask/check though.

Goddamn that's an amazing little laptop,  MBP 14" who?
 
I even like that there's no webcam that would get some black tape put over. 

My HP laptop has SK Hynix 256GB M.2 NVMe SSD that's literally a "bigger" PCB with nothing on it but what Kioxia offers. It's also DRAM-less and given I couldn't find it either separately, integrated SSD controller. Hynix could just as well shrink it to this form factor, but it was apparently not convenient for them to do so given Linus had problems finding anything to stick it in. And yes, that half empty PCB does look funny.
 
This is what it looks like (not my image coz I didn't photograph it). It's so funny how a sticker has it's own spot on an empty PCB as there is so much emptiness.
 
I still don't understand why Toshiba, one of most well known brand names in the world renamed this division to "Kioxia". Why not Toshiba Memory or just remain Toshiba Memory Corporation. I don't get it. It's literally like renaming Ferrari to Ravioli. From well known brand to some pasta. Totally unrelated, totally unknown in a car segment...

I've setup the 100G over Fabric Showcase on thursday for the Amsterdam OCP event next week.
Yes that's crazy (from a normal standpoint) stuff.
It's running a AIC FB127-LX with 3x100G + MZ4LB3T8HALS-00003 (Samsing 3.8TB M.3 NVMe SSDs), a Mellanox MSN2700-CS2F (32x100G) Switch and 4 Compute nods with 100G each.
But Linus, for your "Office use" you should only should get connectx-3 with a MSX1012B-2BRS Switch, it's more cost effective at the moment.
For conclusion some Hardware p*rn:

THIS. I don't understand how using Synology's Quick Connect feature would give Jake access to Linus's files... I own a DS918+ and that's not how this works at all. Quick Connect is just Synology's method of connecting to your NAS using (what I believe is) a reverse proxy, so that you can still connect if your router's Port Forwarding or UPnP shits the bed. It does not automagically grant access to files on the NAS.
 
What they should have done was indeed created separate users on each others' NAS' with their permissions set to only allow access to folders they require. A good way to organize this is to create a share just for backups (for example, a "_backup" share), then create user folders inside said share for each user who needs to backup. Once these folders are created, you just assign access rights for each folder to the corresponding user so they can only access it, and you're good to go.
Yes, you absolutely can run a web server or other form of Linux based server from a Synology - in fact, Synology OS is just a customized flavor of Unix/Linux with a ton of heavily skinned & customized interfaces to reduce the amount of manual configuration one otherwise would have to do. There's also a plethora of 3rd party repositories, allowing you to install many apps Synology doesn't support right from the same package manager.
https://synocommunity.com/packages
https://search.synopackage.com/sources

Jake types in his user/pw on a external system to give linus access to his share - please do not, the other person could catch your credentials.
Create a seperate share & user for the person, who want to backup, also in the share config you can create quotas so your mate don't fill your nas full with his crap.
I upload by "external" Backup to work, where I have racked my own server (verry old system).
But at some friends, I've a Rasip with a usb hdd that openes a reverse SSH tunnel to me (it also could be a cheap vm) and mapps the ssh port there.
Hyper Backup is able to do packupo over ssh, just use "rsync" as backup type. May you have to edit the parth to get it working, in my case I mount the hdd to "/home/$USER/backup". Even on some Webdav services that dont run out of the box just type in "/home" then you can run your backup.

I would use it as a Storage or VM Box, but as storage Box I would use the SAS port together with an expander backplane of a JBOD and put a NIC inside the PCI slot / use a 8x/8x riser for NIC & HBA Card.

From an IT/Sysadmin view, did anyone try Citrix or MS Remote Desktop with iPadOS and a mouse?
If yes, I could start handout iPad's instead of notebooks that have way more problems with drivers etc. and my user hate it, because we have spend to much time with fixing and save about 40% of our first & second level support staff worktime with just go to reinstall the device.

I've setup the 100G over Fabric Showcase on thursday for the Amsterdam OCP event next week.
Yes that's crazy (from a normal standpoint) stuff.
It's running a AIC FB127-LX with 3x100G + MZ4LB3T8HALS-00003 (Samsing 3.8TB M.3 NVMe SSDs), a Mellanox MSN2700-CS2F (32x100G) Switch and 4 Compute nods with 100G each.
But Linus, for your "Office use" you should only should get connectx-3 with a MSX1012B-2BRS Switch, it's more cost effective at the moment.
For conclusion some Hardware p*rn:

Latency is not Round trip time... latency is always one way and not back again.
Yes the ciena "CPE" will add latency, it's mostly a DWDM device without a optical spitter. This is like a switch with "extendes monitoring funktion".
Overhead is a thing, if you test with only 64bytes you will never get the full speed. This is why "more carrier grade" routers like the edgerouter infinity is telling the max. amount of packages at this small size. If you go to 1024-1514bytes you see "mostly the full performance" but you still have overhead, especially if your ISP is limiting your MTU zu 1492 in xDSL or sometimes even smaller on coax lines. Sometimes it's because your ISP packs your IP/Ethernet into SDH or put another header like PPPoE ontop.

The max. speed you will see if you go higher like in a SAN where you use 9000bytes jumbo frames.
It's also possible to configure something larger than 9000 like 12394, so you can add ISP overhead without your customer get's any down sides. Like if you go with Jumbo frames (from customer) and transport it over a link where your command and routing like GRE, MPLS, IPSec adds more overhead, sometimes even combined.
In my sub AS I allow routing of "oversized packets" because my uplink AS accept this.
Then at the next hops:
- some of the IX (Internet Exchange) allow also jumbo, other not so if the connected ISP's can handel it, you have less overhead
- some of your direkt peerings, also allow them like on a link to another compete DC provider (tow customers even move around 30-60GBE to two other DC providers where they are also customers and for whatever reason they decide to go though the "internet" and not via a privat wave)
- even one of our four transit providers allow large packages
 
Testing with a browser is not very efficient, to much overhead like http, encryption, etc.
 
Peering at a IX or direkt is cheaper than transit. Transit is just someone is doing peering all over the world for you. This is why you have a 10G Wave to the DC and all connected peering partners (who have 10g or more) but "only" 5.xxxG to the "internet".
 
And then the layer 0 & 1, the distance on the SFP is only a slide tast, how good it could perform. The only real thing is the power budged, some 10km SFP's have a budget of 8,2db and the customer is using it on a 18,8km link...

Glad you guys got it up.
 
Few points:
-Dark fiber is actually a term used to describe leased fiber, not unused fiber. Its called dark because you cant see the light on the fiber because its in used by the customer leasing it. 
-Using a LR (long range) SFP with a SR (short range) will not attenuate the light. Whats actually happening is its overheating one of the ends because its receiving too much light and burning the lazer.
- There is no such thing as a cheap SFP.... FS.com.... only overpriced SFPs
 
So all in all you are now part of the ISPs network. We deploy Ciena and that exact model and use it for layer 1 transport (yes its actually layer 1).  That 6500 they are using is common in transport and they are actually on a 10gig wavelength off a 100gig XFP on the main Ciena chassis. 

Im probably going to be a 'brave bastard' then
 
To be fair though, if ur a pure gamer, so long as u dont use your PC for anything truly important, viruses are not really that much of a worry especialy if ur not browsing dodgy websites.
If u really must use ur gaming PC to browse 'those sites' , use a removable SSD with a W7 installation on it to hot swap in purely for such browsing sessions.
If/when viruses infect ur system, just format and start over using a backup image, quick and simple.
 
Honestly, for a purely gaming system, the worry about security is marginal, its unlikely , highly unlikely, ull get infected, and if u do u shouldnt have anything ur worried about loosing on ur system anyway, so just format.
 
I will consider upgrading to W10 when W7's time comes, but only if theres a stripped down pre modded, trusted version available with all the crap removed, and full control available to the user.
I have no intention of using W10 in its current state, and i really cba to go through the hassle of modding W10 myself using the various programs and scripts currently available, i shouldnt have to do that, and i wont whilst my current W7 installation is perfectly capable.

Ok you told you want to go the 100G way. I've a demo Setup in my Work Lab with a Mellanox MSN2700-CS2F & MCX416A-CCAT and some QSFP28 DAC's running well. Main Target of this setup is to show the 300G NVMe Storage box underneath it (delivered by memorysolution).
I would not do VM Storage on the Optane Drive, because if you have (dual or quad) 100GbE you will not have the need for local drive, because the latency will be unfeelable low.
For Passthrough - I recommend to forward one USB Controller to each VM so any device will be on the VM when plugged in the "right" card.

I still want to see the AMD version, surely they'd want to send over some EPYC stuff to compete and compare? Maybe when 7nm EPYC is out?

I've tested it here in germany with a low latency connection (about 9ms to the Host's public IP via our own AS  -> Telia -> cogent -> blade), a normal Coax uplink (150/10 40-60ms unitymedia -> cogent -> blade) & a VDSL (250/40 22ms german telekom -> hupus -> blade) uplink.
Also tested Mobile connection on Train.... fail.
The wired uplinks worked fine, I've didn't notice any "lagging".
Only downside is, that my w520 & x201 GPU's does not like decoding of the stream, because I can hear the sound of the Hitman shooting, but had a decoding delay of about 300-800ms. So I'm more into this oldschool type of computing and because of that the GPU's Video Engine can't get the performance needed.
Alternativly I also test it on my old AMD Phenom 9750 + ATI HD5770 and on my NUC with i5-5250u, then it works fine.
Also the Steam link was able to run gread with a small hack - Setup a VPN endpoint on your home firewall, configure the VPN target on the shadow system, establish connection, tell the steam link the VPN IP of the shadow... works for me.
And because Valve stop selling their link, the ghost will be an alternative as soon it gets buyable. Hope they bring dual or 3 monitor clients soon. Because I use this as a working setup. In my Main house (weekend) & my work flat I've a fixed monitor setup that I can't carry around 3 times a week and having local computers also does not make any sense. Yes I have laptops, but because of my needs, they can't be newer then 2***/3*** gen cpu's (legacy bios needed).

I've tested it here in germany with a low latency connection (about 9ms to the Host's public IP via our own AS  -> Telia -> cogent -> blade), a normal Coax uplink (150/10 40-60ms unitymedia -> cogent -> blade) & a VDSL (250/40 22ms german telekom -> hupus -> blade) uplink.
Also tested Mobile connection on Train.... fail.
The wired uplinks worked fine, I've didn't notice any "lagging".
Only downside is, that my w520 & x201 GPU's does not like decoding of the stream, because I can hear the sound of the Hitman shooting, but had a decoding delay of about 300-800ms. So I'm more into this oldschool type of computing and because of that the GPU's Video Engine can't get the performance needed.
Alternativly I also test it on my old AMD Phenom 9750 + ATI HD5770 and on my NUC with i5-5250u, then it works fine.
Also the Steam link was able to run gread with a small hack - Setup a VPN endpoint on your home firewall, configure the VPN target on the shadow system, establish connection, tell the steam link the VPN IP of the shadow... works for me.
And because Valve stop selling their link, the ghost will be an alternative as soon it gets buyable. Hope they bring dual or 3 monitor clients soon. Because I use this as a working setup. In my Main house (weekend) & my work flat I've a fixed monitor setup that I can't carry around 3 times a week and having local computers also does not make any sense. Yes I have laptops, but because of my needs, they can't be newer then 2***/3*** gen cpu's (legacy bios needed).

@Egg-Roll If you calculate the systems livetime with 5 years and you use it a lot for gaming ok. Also I'm with you with storage needs and a bit on the Uplink pricing (because all places I go to have good connectivity).
But if you need this kind of resources on the go / multi homed.. you would buy more systems?

I've tested it here in germany with a low latency connection (about 9ms to the Host's public IP via our own AS  -> Telia -> cogent -> blade), a normal Coax uplink (150/10 40-60ms unitymedia -> cogent -> blade) & a VDSL (250/40 22ms german telekom -> hupus -> blade) uplink.
Also tested Mobile connection on Train.... fail.
The wired uplinks worked fine, I've didn't notice any "lagging".
Only downside is, that my w520 & x201 GPU's does not like decoding of the stream, because I can hear the sound of the Hitman shooting, but had a decoding delay of about 300-800ms. So I'm more into this oldschool type of computing and because of that the GPU's Video Engine can't get the performance needed.
Alternativly I also test it on my old AMD Phenom 9750 + ATI HD5770 and on my NUC with i5-5250u, then it works fine.
Also the Steam link was able to run gread with a small hack - Setup a VPN endpoint on your home firewall, configure the VPN target on the shadow system, establish connection, tell the steam link the VPN IP of the shadow... works for me.
And because Valve stop selling their link, the ghost will be an alternative as soon it gets buyable. Hope they bring dual or 3 monitor clients soon. Because I use this as a working setup. In my Main house (weekend) & my work flat I've a fixed monitor setup that I can't carry around 3 times a week and having local computers also does not make any sense. Yes I have laptops, but because of my needs, they can't be newer then 2***/3*** gen cpu's (legacy bios needed).

agree to disagree...
this is NOT going to happen by multiple factors, like:
how you wanna pull HDMI quality (18Gbps) thru AT&T/Verizon infrastructure?
why x265 when you are forced to send 60+ I-frames only anyway?
how is it better/acceptable to watch 60 wash-out JPG over crystal clear RAW render?
anyone wanna bet that gaming PC+streaming PC rent will EVER get cheaper than just owning one anyway?
so they run a server; that runs VM; that runs on VM; that runs denuvo VMprotect; that is on VM... clever...
---
This project is already dead as every predecessor. ..., but if we are lucky, in few years time, we'll get a linux open-source code for screen-streaming over your home network.

I've setup the 100G over Fabric Showcase on thursday for the Amsterdam OCP event next week.
Yes that's crazy (from a normal standpoint) stuff.
It's running a AIC FB127-LX with 3x100G + MZ4LB3T8HALS-00003 (Samsing 3.8TB M.3 NVMe SSDs), a Mellanox MSN2700-CS2F (32x100G) Switch and 4 Compute nods with 100G each.
But Linus, for your "Office use" you should only should get connectx-3 with a MSX1012B-2BRS Switch, it's more cost effective at the moment.
For conclusion some Hardware p*rn:

I've setup the 100G over Fabric Showcase on thursday for the Amsterdam OCP event next week.
Yes that's crazy (from a normal standpoint) stuff.
It's running a AIC FB127-LX with 3x100G + MZ4LB3T8HALS-00003 (Samsing 3.8TB M.3 NVMe SSDs), a Mellanox MSN2700-CS2F (32x100G) Switch and 4 Compute nods with 100G each.
But Linus, for your "Office use" you should only should get connectx-3 with a MSX1012B-2BRS Switch, it's more cost effective at the moment.
For conclusion some Hardware p*rn:

I work at a Datacenter / multi customer for two years, but I can't see any decrease of dual socket systems.
Quad socket systems are decreasing, because of the better density. Especially if they can live with 1,5TB RAM or less.
Singe CPU Systems are also getting less, because most of the customers want as much computing power in their racks as possible.
But I can see a increase of single cpu on "non standard compute workloads" like backbox firewalls like Fortigate's, VPN Gateways like JunOS Puls gateways or other networking or transfer intensive loads. Mostly because the need to get more efficient.