unevent

Seasonic and those made by Seasonic is my recommendation. If you really need 750W, Newegg has some sales and rebates on several that are that size, but they are Gold vs. Titanium. Assuming you are in the US, Seasonic PRIME Ultra 750W 80+ Gold, fully modular, 12 year warranty, $109.99 (sale ends this coming Sunday), $20 rebate brings final to $89.99 (free shipping). The same in Titanium would be $180. I don't believe you would ever realize the ROI with that price difference as efficiency is only around 1.5% better with Titanium vs. Gold.

Pegged out half the cores according to the GUI? Was eight cores detected and eight prime95 threads started? If it is just the GUI then may be already discussed elsewhere as @1812 mentioned.

Post a diagnostics as well so we can see what is being found on boot. GUI - Tools - Diagnostics.

You can download Linux 64bit Mersenne Prime (Prime95) and it will load all cores to 100%. Extract to a folder on unRAID, telnet/SSH to the folder and use the commands below. chmod +x mprime 'This will make it executable ./mprime 'To have a choice of test options and number fo cores to use ./mprime -t 'or use this to jump into torture test on all cores without selecting any options This will load all available cores unless you specify less than actual manually.

My opinion is the goal is to increase capacity of spinner drives while not negatively affecting the performance. Going to multi-actuator and keeping the costs and reliability in check is the first step given current tech. But yes, splitting a drive into a RAID0 moves you to SSD-like performance, and there may be a market for that if NAND prices continue to be astronomical.

In theory this dual actuator will increase performance because of the ability to read or write on two heads at same time due to two independent sets of heads. It would be up to the drive firmware and processor to realize any further improvements.

Not really going to get pure sine in that price range for rack. Stepped/PWM is it. If you want pure sine then suggest going with one of the smaller tower 'shoebox' units from APC, TrippLite, or CyberPower. CyberPower CP900EPFCLCD 900VA/540W, 148€

The Tripplite has C14 input, 7 C13 outlets, 2 C13 to C14 cables. I suppose an adapter or stick with the tower units. Not many options in low capacity rack mount. From Triplite: UPS Input Connection Description - C14 inlet accepts user-supplied AC power cord with country-specific plug. Adapter

Rackmount options: APC SC450RMI1U 450VA/260W, 1U, 213€ TrippLite SMX500RT1U 500VA/300W, 1U, 239€ For tower options, CyberPower, APC, TrippLite in similar output range are 35-50% of the above. Any of those brands will work with apcupsd or NUT on unRAID.

Have you measured your watt load? Limiting to <100€ pretty much excludes rack mount.

Suggest buying an inexpensive Ethernet switch to give you more ports.

Thanks, and nice work.

Is this going to be 'the new way' of doing things or just one possible way for 6.4? I have no need or desire to segregate Dockers from VMs using VLANs and then juggle with rules on my router when I need communication between VLAN. I do have a need and desire to be able to give each Docker and VM there own IP and MAC, which I can do now just fine in 6.3.5 and on same subnet.

Yes! Good to mention. The alias created to route select clients thru VPN should also be added to the exclude list in Squid config to have those also bypass the proxy and should exclude the VPN interface from the Squid bind list. Edit: Can also get DNS leaks by having DNS queries go out the WAN interface vs. VPN interface, so don't forget to configure DNS interface and rules accordingly.

Caching proxies are really not as beneficial these days given the large Internet pipes, even your 20/1. They slow the Internet experience because you are constantly working from the cache by writing to/checking if exist/reading from disk, even with SSD. The Squid package in pfsense is what you would use for the transparent proxy which will proxy non-encrypted traffic. Not much these days is non-encrypted so benefit varies. You can do encrypted traffic cache proxy by configuring certificates you install on all client devices, but headache if you have many devices. There is another way to do encrypted caching without certificates, but can give some browsers/devices fits. It used to be good for caching Windows update stuff, but MS changes delivery and becomes unreliable and you are always having to tweak the filters to capture the updates to cache. The QoS is no where near as simple to configure as one would have experienced with Toastman Tomato. There are books, videos, etc. on it though. Pi hole (ad blocking) can be done with pfblockerNG and works extremely well, add Snort for IDS/IPS.

Most of my containers are from Binhex, though I don't use any of the VPN wrapped which was why I asked the question. I've found them to be the most stable and not crazy-fanatical about updating on what seems like every breath as some others do. Which reminds me, need to send Christmas gift to put some kick in the eggnog.

Very good, thanks.

Curious if you are routing DNS queries through the VPN tunnel or in the clear to Google servers?

Picked up early present the other day, Ryzen 7 1700 and ASRock X370 Taichi to replace ancient Phenom II X4 940.

Ahh, Gold membership...well worth the money, IMO, if need that kind of support. Edit: Some of their forum members can be intimidating, but there is a lot of expertise available there and most likely any question outside of something completely unique has already been answered.

$99 to access the manual? You mean the subscription for paid support? There are multiple books available (even at your local library) for pfsense and their forum is usually all you ever need. It is free software, they have to pay the bills somehow.

Assuming you have nfs-common installed and an fstab entry already made, just add the following to the fstab entry using your user name and password: username=santa,password=claus

Late, but was bored. Font is based on Bank Gothic with SpaceX inspiration.

I've never tried it with Powerchute connected via USB then try to slave apcupsd to Powerchute that way. Typically use a UPS management card with Powerchute then configure apcupsd using pcnet pointed to the powerchute as client. It may work same way, but not sure. Configure Powerchute software to have unRAID IP as client and set name/pass then configure apcupsd for pcnet and name/pass. Apcupsd should connect with Powerchute over network and Powerchute should talk to the UPS via USB if it works. Assuming the Windows 10 box is to be master (the last one to turn off).

The sabnzbd switches don't work if the Docker container does not have nice, ionice, and the multi-core par2 binary.