Plex Media Server shutting down my router(s)?

General Discussions
1

Hi,
For the last 4-5 months I have had my cable modem shutting down internet connection regularily.

I have been error seeking for a while, and before I list some of the initiatives I and my internet provider have done, here´s why I am writing now in the Plex Forums: A pattern that is more than a coincidence is that when the internet shuts down, it is often when friends outside my network stream from my Plex Media Server. They use Nvidia Shields, Apple TVs, Chromecasts, Samsung TVs and more.

My internet connection goes down, both wired and wireless, and I have to restart the cable modem.

  • I have a dedicated IP-address for Plex Media Server Windows PC and portforwarded 32400 in my router
  • I have asked my friends to log out and login in again on their Plex apps
  • I have had the cable modem replaced from the internet provider with a new unit - and also a new model.
  • I have the latest firmware in the cable modems / router
  • I had a technician from the internet provider remotely check the quality of the signal
  • I had a technician from the internet provider physically check the quality of the COAX signal, replace wall sockets, replace cable connectors, in my home and in the infrastructure outside
  • I have removed my own bridged router from the setup, now only use cable model from internet provider
  • Plex Media Server is registered in antivirus software

One would think that software shouldn´t be able to shut down internet connection. Wonder if Plex has some traffic that the cable modems/routers don´t like?

Usually @sa2000 and @OttoKerner have interest in cases like this.

SIMILAR ISSUES (some are very old)

Keep losing internet connection

Plex app causes router to reset
https://forums.plex.tv/discussion/40454/plex-app-causes-router-to-reset

Is Plex Media Server crashing my Connection to the Internet?

Plex crashes ‘no internet access’ randomly

Plex frequently and consistently crashes wifi network

MY SETUP

Plex Media Server:
NAS [Synology DS415+] + 3 x 6 TB HDD [WD60EFRX], Synology Hybrid Raid, SHR
HP ENVY 15-j160no Notebook PC
Windows® 10 64 Bit
Intel® Core™ i7-4700MQ, 2.4 GHz, 6 MB cache, 4 cores, CPU Benchmark 7,827
16 GB RAM
250 Gb Samsung SSD 840 Evo (160 Gb free)
NVIDIA GeForce GT 840M GPU (2 GB dedicated, DDR3)
Norton 360 Antivirus
Plex Media Server [version 1.2.2.2857]
PMS is only usage of this laptop/server

Local setup:
Internet: 300Mbit/60Mbit
Cable Modem [now SagemCom F@st 3686/V2 previously Netgear C6250EMR]
Netgear Nighthawk R7000 Smart WiFi Router, 802.11ac (prev. bridged, removed for now)
Nvidia Shield Android TV
Denon Surround Receiver [AVR-3808]
Philips TV [55PFL8008S]
Cat 7 Ethernet cables (except Nvidia Shield)

3

When I have had issues like this crop up with friend’s setups, I almost always have the ISP put the modem into bridge mode and put a prosumer or high quality consumer router as the first part of the LAN.

In fact, this is exactly what I did on my own setup. The modem provided to me wouldn’t allow port forwards on 32400. I could make the forward, but the logs showed that this port was locked out in firmware with the error messages. As soon as I put the modem into bridge mode, I used my own router and set my port forwards on that, and everything works fine. In bridge mode everything is passed through and the ports aren’t looked at by the device itself. It doesn’t care at that point.

This looks like it’s an issue with the modem itself, or the firmware running on it.

Your router is one I would consider a higher end consumer/prosumer model and can easily handle the traffic. So it would have the external IP on it’s WAN port, not your modem (which is suspect at this point.) The router handles firewall, port forwards, DHCP, etc. It becomes the main active connection to the internet, not the modem.

It looks to me as though you had the wrong device bridged.

EDIT: The Netgear C6250EMR can easily be put into bridge mode. http://www.downloads.netgear.com/files/GDC/R6250/R6250_UM_13Apr2015.pdf on page 95 tells you how to do it. After that, restart your NightHawk and verify it has an external IP on the WAN port, and set up your port forwards. Check each on http://canyouseeme.org. (Verify the external IP there as well.)

EDIT again: DON’T rely on uPnP for handling your ports. Even higher end routers can (and do) clear those ports with no activity on them, resulting in your connection disappearing randomly. (The number 1 cause of intermittent Remote Access.) Forward ports on the router. It is always ready for traffic then.

4

Thanks @trumpy81 and @MikeG6.5,

3 weeks ago I received the “new” SagemCom F@st 3686/V2 as replacement for Netgear C6250EMR from the internet provider. Both were set to Bridge mode, while my primary router has been Netgear Nighthawk R7000 since February 2014. Yes, 2,5 years.

The issue persists with both SagemCom F@st 3686/V2 and Netgear C6250EMR - 2 models - 2 different brands.

In relation to the cable internet provider, they/we have systematically error searched (on location measurements, logging, changed wall plugs, replaced connectors, new cable modem, removed my own “unauthorized” router from the setup).

UPnP is enabled
DHCP Lease time 3600 seconds
Static address for the PMS server, portforwarded 32400

I have had static IPs for most of my equipment. The “new” SagemCom F@st 3686/V2 can only handle 16, which is actually not enough. I have to prioritize then. PMS server no. 1.

My own theory is the COAX infrastructure itself. My internet provider has stretched the performance recent years. Early 2014 I had 60/10 Mbit. Now I have 300/60Mbit - and later this year I can order 1Gbit. Firmware is also an unexplored error source. I am also wondering if the DHCP Lease time, which looks very short to me, could make my cable model go into limbo.

6

Along the lines of power problems, get a small UPS and use that on the modem. So that way you don’t lose the internet connection right away when the power should fail. And a fairly good one will also act as a line conditioner. Or give you warnings should the power fluctuate outside the norm.

7

Thanks again @trumpy81 and @MikeG6.5,

My next step is to talk DHCP Lease time 3600 seconds with the cable provider. They might have advantage of the input. 1 hour intervals appear way too short for private clients - if it had been a cafe, hotel or other public space, I could understand. Feedback for their firmware.

I will also find out what this setting was on the previous cable modem C6250EMR, and my personal Netgear R7000. Googling R7000 it appears that DHCP lease time to be hardcoded, but standard seems to be 86,400 seconds (24H).

I´ll check the AC powerlines, e.g. separate the cable modem to its own alternate power outlet.

I have considered connecting a “Smart Plug” that cuts the power and resets the cable modem daily on a convenient time.

During my recent research I found that my cable provider is testing a brandnew DOCSIS 3.1 model Sagemcom F@st 3890. Maybe I should try to get hands on that.

EDIT: I have changed DHCP Lease time from 3,600 to 129,600 Sec (36 hrs). This may reduce the probability for the IP-config to mess up, but will eventually happen.

9

Have you investigated the cable line coming into your house? I had a very similar issue, where my internet would drop/disconnect occasionally (and I only could really notice when there was internet activity) and couldn’t figure it out for the life of me. After two years of trying to figure it out, I eventually found that it was the cable line running from the cable company’s pedestal to my house. Techs had checked the signal quality and replaced the demarc box and all connections inside and it turned out to be the damned drop.

10

Hi again @trumpy81 @MikeG6.5 @interconnect
Thought I´d give a heads up for this issue, since I have reasons to believe it´s solved and may be useful for others.

Since October I have systematically error searched my network. Big changes - on top of what I mentioned in OP - were removing my Bridged Router (no help), remove set top TV box from Cable provider (no help), extend the DHCP Lease time from 3600 seconds to 2592000 seconds/30 days (no help).

An element in my network I had missed was a TP-Link Powerline (TL-PA8030P). It was plugged into my power circuit, but was not part of the Plex setup (NAS, laptop, Shield…). Browsing through the Router logs indicated that the Router break downs happened when I was Plex streaming external = Uploading activity at the same time as when the powerplugs renewed IP address for attached device. Powerplugs establish a separate invisible network.

Mid December I upgraded to newest firmware for the 2 TL-PA8030P Powerlines, and I have only had to restart my router once in recent 5 weeks! So far so good…

HAPPY NEW YEAR! :-bd

11

Good to hear you solved it! I would recommend though that you take the time and run Cat6 Ethernet cable instead instead of using the power line adapters. You’ll be much happier in the end.