wayner

I doubt even that scenario is as bleak as you portray. I am a long-time user of SageTV. SageTV was purchased by Google as they used some of the software in the TV part of their Google Fiber business. But SageTV as a separate product went dormant, but there was (and still is) a vibrant user community. A couple of years ago when Google wasn't using SageTV as much the founder of SageTV had Google Open Source SageTV. So SageTV V9 is now an open source product (and runs very well under unRAID in a Docker). I could see something like that happen with unRAID if it were bought by a large company and then ignored.

I just installed the PiHole docker yesterday and so far it is awesome. Thanks to those who wrote the software and the docker!

Me too on Unifi - I have a USG plus five APs. Are you sure that you need a 10Gig network? Do you know if your components can take advantage of that much speed? If you have fast PCs they might but I doubt anything else will need that much speed - even 100Mbps would be more than enough for printers, TVs, streaming boxes, etc. Wifi devices will obviously not get anywhere near those speeds in the real world. The bottleneck might be somewhere other than the network and then you have needless spent a lot of money.

And if it is generating that much heat then it is probably burning a lot of electricity. If that is the case you might be better off getting a NUC PC and hide it away in the office - you can get pretty powerful units as well. I kind of like the idea of having everything in one box in some ways, but then you really have all your eggs in one basket. If your system goes down then you are really stuck.

In many ways I agree with you, but HDMI over Ethernet is very popular in high-end centralized video distribution which is often part of the Home Automation system in high-end homes. I have a Control4 Home Automation system and while I do not have centralized video many of the folks who use Control4 do have these types of systems and spend tens of thousands to have them installed. Systems like Leaf, Atlona and Just Add Power are three examples of this. The reason why I say this AV Access system looks too good to be true is that the Atlona Tx-Rx pair will set you back $725. That's almost 5X the price of the AV Access product.

Thanks Hoopster. Is the 10 bit support needed for HDR content?

If you have other Cat6 runs you can use USB over ethernet adapters that are pretty cheap. But these need a dedicated cable - you can't use a switch. You could always use your phone as your remote. Or and old phone or tablet.

For Librelec you may not need the USB as you can likely use some for of IP control. That AV Access system looks like it may work but that is more for AV than for PCs. The issue might be getting USB up to your office unless you can use the RS232 serial ports for mouse and keyboard. The RS232 would generally be used for serial control of a TV or AVR. It does not support the better HDR formats but that may not matter for you. The only issue is that the price looks too good to be true. I would have expected something like this to be at least $500 if it supports 4K. But it is on Amazon and it is pretty painless to send it back if it doesn't work.

Sorry - I guess it is just built into the core system no need for a plugin. And I would highly recommend a UPS for any and all servers that you would have.

MS is moving forward on a few things to make Windows more Linux-friendly. In addition to the Bash prompt the latest edition has beta versions of ssh - both client and server. This ssh server seems to have lots of issues, but ssh client works very well and negates the need for using Putty. And you can set up auth keys to facilitate logins. Mind you this is less needed with unRAID since we now have the terminal within the web UI.

Thanks - I wasn't sure if Skylake had HEVC support and from what you are saying it does not.

One way to get more used to Unix is to use the Ubuntu Bash sub-system for Windows. You can install this on Windows 10 and it can do a lot of the things that you can do in Linux in a terminal window. It can be very useful for running code that is written for Linux in languages like Python. It is kind of like Linux with Windows training wheels.

Products like Just Add Power and other video Matrix switches form the heart of centralized video distribution. The folks that have these swear by these but they can be pretty expensive if you have a lot of TVs - think $5-10k or more. Just Add Power is on the cheaper end but if you want 4K it gets more expensive. And pretty much all of these systems need a dedicated Ethernet run from your central rack to each room where you are distributing power. Then you may want a HA system like Crestron or Control4 to control it all. And then you may also want to think about centralized Audio distribution with ceiling speakers, matrix amp, etc. If you still want centralized video an alternative to consider is Video Storm Net Play. It uses AndroidTV devices like FireTVs or ShieldTVs at each TV as a receiving device. But you can put almost anything into it using their Encoders. This also works well for stuff like Video Tiling/PIP or Video Walls. Depending on what you want to do you could probably get away with spending $2--3k. But then the logical question is: If I am putting a device like an Nvidia Shield at every TV then why can't I use that as my source for everything? The thing that may be hardest to replace right now is cable TV and a centralized system like this can help you reduce the number of cable STBs that you rent. For your last question - that is possible but then you may need lots of HDMI cabling through your house. You may also need a way to get IR or USB connections over the wiring. You may also need to run several VMs and have multiple video cards. There are threads here on people who have done systems like this.

Another vote for Unifi USG here. I have five APs in and around my house plus the USG and I will likely be adding Unifi switches in the future. I run the Unifi controller in a Docker container on my unRAID server.

HDMI to Ethernet baluns? You may also need to transmit USB over the Ethernet as well. Going 4K will make this a lot more expensive. Many of these types of products are often used for centralized video distribution and not so much for PCs - they may introduce some latency.

If you were to buy a new CPU today with a thought towards being able to transcode 4K content then what CPU would you recommend? A Coffee Lake CPU like the i7-8700 or i7-8700K? A Skylake 7800X? Or does Skylake have the GPU capabilities?

And unRAID has a plugin to help manage the UPS.

Time to buy a (bigger) UPS?

Is the data from 2013 or is that just the date of the original blog post?

Thanks Bob - I will try that when I get a chance. Even with this I haven't really been having issues. I did have a crash or two when the system was first built a year ago but it has been pretty stable since then. The system went down one other time on Dec 30 but I am not sure if that was a crash or if something else happened. I was thinking that I didn't have any PCI cards but I think I do have a FireWire card to use for changing channels on my STBs. I think it is the older PCI rather than PCIe is that matters.

The USB is from a Asus P8H67-M Evo Asus motherboard and has an i5-2500K CPU. The USB chipsets are: ASMedia® USB 3.0 controller : 2 x USB 3.1 Gen 1 port(s) (2 at back panel, blue) Intel® H67(B3) chipset : 12 x USB 2.0 port(s) (4 at back panel, black, 8 at mid-board) I am not sure which ports I am using but I am pretty sure it is the USB 2.0 ports. I do have two Hauppauge HD-PVRs plugged in as I use this system as a SageTV DVR/media server with the SageTV docker and the LibreElec DVB edition of unRAID. But I never plug things in or out, but it is possible that the HD-PVRs may go up or done that could cause it to behave as if I new USB device was plugged in.

Did you ever fix this. I get this error message on my system although I am not sure that it is causing any issues. I had this under 6.3.5 and now under 6.4.1. On my monitor it also gives the Disabling IRQ 16 message. Feb 3 19:14:57 Hoylake kernel: irq 16: nobody cared (try booting with the "irqpoll" option) Feb 3 19:14:57 Hoylake kernel: CPU: 0 PID: 0 Comm: swapper/0 Tainted: G C 4.14.16-unRAID #1 Feb 3 19:14:57 Hoylake kernel: Hardware name: System manufacturer System Product Name/P8H67-M EVO, BIOS 3602 10/31/2012 Feb 3 19:14:57 Hoylake kernel: Call Trace: Feb 3 19:14:57 Hoylake kernel: Feb 3 19:14:57 Hoylake kernel: dump_stack+0x5d/0x79 Feb 3 19:14:57 Hoylake kernel: __report_bad_irq+0x32/0xac Feb 3 19:14:57 Hoylake kernel: note_interrupt+0x1d4/0x225 Feb 3 19:14:57 Hoylake kernel: handle_irq_event_percpu+0x39/0x3f Feb 3 19:14:57 Hoylake kernel: handle_irq_event+0x31/0x4f Feb 3 19:14:57 Hoylake kernel: handle_fasteoi_irq+0x8c/0xf3 Feb 3 19:14:57 Hoylake kernel: handle_irq+0x1c/0x1f Feb 3 19:14:57 Hoylake kernel: do_IRQ+0x3b/0xbb Feb 3 19:14:57 Hoylake kernel: common_interrupt+0x98/0x98 Feb 3 19:14:57 Hoylake kernel: RIP: 0010:ip_sabotage_in+0x2b/0x31 Feb 3 19:14:57 Hoylake kernel: RSP: 0018:ffff88061f203a60 EFLAGS: 00000282 ORIG_RAX: ffffffffffffff9e Feb 3 19:14:57 Hoylake kernel: RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000101536f4c Feb 3 19:14:57 Hoylake kernel: RDX: ffff8805fe8a2090 RSI: 00000000fffffe01 RDI: ffffffff8150ba23 Feb 3 19:14:57 Hoylake kernel: RBP: ffff8805637c0e80 R08: 0000000000000001 R09: 0000000000000000 Feb 3 19:14:57 Hoylake kernel: R10: ffff88061f2039a0 R11: ffff880602c150a0 R12: ffff8801107ffae0 Feb 3 19:14:57 Hoylake kernel: R13: ffff88061f203aa0 R14: 0000000000f00008 R15: ffff8801107ffae0 Feb 3 19:14:57 Hoylake kernel: ? ip_finish_output2+0x278/0x2c2 Feb 3 19:14:57 Hoylake kernel: nf_hook_slow+0x37/0x96 Feb 3 19:14:57 Hoylake kernel: ip_rcv+0x2f2/0x346 Feb 3 19:14:57 Hoylake kernel: ? ip_local_deliver_finish+0x1b8/0x1b8 Feb 3 19:14:57 Hoylake kernel: __netif_receive_skb_core+0x6ba/0x733 Feb 3 19:14:57 Hoylake kernel: ? ip_output+0xc4/0xcf Feb 3 19:14:57 Hoylake kernel: netif_receive_skb_internal+0xbb/0xd0 Feb 3 19:14:57 Hoylake kernel: br_pass_frame_up+0x12d/0x13a Feb 3 19:14:57 Hoylake kernel: ? br_port_flags_change+0xf/0xf Feb 3 19:14:57 Hoylake kernel: br_handle_frame_finish+0x429/0x459 Feb 3 19:14:57 Hoylake kernel: ? ipt_do_table+0x55f/0x587 [ip_tables] Feb 3 19:14:57 Hoylake kernel: ? br_pass_frame_up+0x13a/0x13a Feb 3 19:14:57 Hoylake kernel: br_nf_hook_thresh+0x93/0x9e Feb 3 19:14:57 Hoylake kernel: ? br_pass_frame_up+0x13a/0x13a Feb 3 19:14:57 Hoylake kernel: br_nf_pre_routing_finish+0x268/0x27a Feb 3 19:14:57 Hoylake kernel: ? br_pass_frame_up+0x13a/0x13a Feb 3 19:14:57 Hoylake kernel: ? nf_nat_ipv4_in+0x21/0x68 [nf_nat_ipv4] Feb 3 19:14:57 Hoylake kernel: br_nf_pre_routing+0x2d8/0x2e8 Feb 3 19:14:57 Hoylake kernel: ? br_nf_forward_ip+0x32c/0x32c Feb 3 19:14:57 Hoylake kernel: nf_hook_slow+0x37/0x96 Feb 3 19:14:57 Hoylake kernel: br_handle_frame+0x2a0/0x2d3 Feb 3 19:14:57 Hoylake kernel: ? br_pass_frame_up+0x13a/0x13a Feb 3 19:14:57 Hoylake kernel: ? br_handle_local_finish+0x31/0x31 Feb 3 19:14:57 Hoylake kernel: __netif_receive_skb_core+0x463/0x733 Feb 3 19:14:57 Hoylake kernel: ? napi_gro_receive+0x42/0x76 Feb 3 19:14:57 Hoylake kernel: ? rtl8169_poll+0x49e/0x4bc [r8169] Feb 3 19:14:57 Hoylake kernel: process_backlog+0x8c/0x12d Feb 3 19:14:57 Hoylake kernel: net_rx_action+0xfb/0x24f Feb 3 19:14:57 Hoylake kernel: __do_softirq+0xcd/0x1c2 Feb 3 19:14:57 Hoylake kernel: irq_exit+0x4f/0x8e Feb 3 19:14:57 Hoylake kernel: do_IRQ+0xa5/0xbb Feb 3 19:14:57 Hoylake kernel: common_interrupt+0x98/0x98 Feb 3 19:14:57 Hoylake kernel: Feb 3 19:14:57 Hoylake kernel: RIP: 0010:cpuidle_enter_state+0xe0/0x135 Feb 3 19:14:57 Hoylake kernel: RSP: 0018:ffffffff81c03ec8 EFLAGS: 00000246 ORIG_RAX: ffffffffffffffad Feb 3 19:14:57 Hoylake kernel: RAX: ffff88061f220900 RBX: 0000000000000000 RCX: 000000000000001f Feb 3 19:14:57 Hoylake kernel: RDX: 000014814bafc212 RSI: 0000000000020140 RDI: 0000000000000000 Feb 3 19:14:57 Hoylake kernel: RBP: ffff88061f228600 R08: 000043fc5a01fd81 R09: 0000000000000018 Feb 3 19:14:57 Hoylake kernel: R10: ffffffff81c03ea8 R11: 0000000000000048 R12: 0000000000000001 Feb 3 19:14:57 Hoylake kernel: R13: 000014814bafc212 R14: ffffffff81c59138 R15: 000014814baf9ec5 Feb 3 19:14:57 Hoylake kernel: ? cpuidle_enter_state+0xbb/0x135 Feb 3 19:14:57 Hoylake kernel: do_idle+0x11a/0x179 Feb 3 19:14:57 Hoylake kernel: cpu_startup_entry+0x18/0x1a Feb 3 19:14:57 Hoylake kernel: start_kernel+0x3ea/0x3f2 Feb 3 19:14:57 Hoylake kernel: secondary_startup_64+0xa5/0xb0 Feb 3 19:14:57 Hoylake kernel: handlers: Feb 3 19:14:57 Hoylake kernel: [] usb_hcd_irq Feb 3 19:14:57 Hoylake kernel: Disabling IRQ #16

I have it in bridge mode but I also had to forward UDP on port 3478 to deal with the STUN error. But that may now be fixed in the docker template. That said - when I upgraded to 6.4.1 on Saturday and update the firmware on my UAPs I noticed some STUN errors but they seemed to go away on their own. I may try switching to br0 mode and giving it its own IP address as that may negate the need to worry about ports as much. And then you will have fewer worries about port conflicts.

I think it is fine the way it is, except I hate where there is support for multiple dockers that have nothing to do with each other in one thread.

Does anyone else find that this plugin provides inconsistent results that are nowhere near what you get when you do a "manual" Speedtest? My "rated" service is 250/20. When I run speedtests by going to the web page I generally get results of approximately 320/21. But with this plugin the results are WAY lower. I run this daily at 6am when there should be very little traffic on my LAN or local node. My download speeds over the last few weeks have ranged from 22 to 284 Mbps. The average is 167Mpbs. That just doesn't seem right.