recently we moved houses and are now using VDSL for our internet connection. Previously we had Cable and before that I was working with the same Readynas on Glassfibre internet as well.
Now with VDSL we have a AVM 7530 Fritzbox and I am seeing some strange interactions.
The moment I connect the Readynas to the modem, the internet connection goes down. I can see in the Fritzbox error log that the connection is interupted. The DSL connection is still intact, but I get a PPPoE fault. 30 seconds later the connection is briefly established, after which is is just as quickly interupted again.
The moment I disconnect the Readynas, the connection is fine again. In the modem I have now limited the internetaccess of the Readynas and this as well leaves the internet connection intact even while the Readynas is still physically connected.
It seems there is some external communication responsible for this behaviour. At this moment I can still access the NAS internally on the LAN, but the device itself no longer has access to the WAN (or the other way around).
Does anyone know what could be causing this and how to resolve this?
Model: RN31200|ReadyNAS 300 Series 2- Bay (Diskless)
Does anyone know what could be causing this and how to resolve this?
If the ReadyNAS MTU is set to 1500, try lowering it to 1492.
That seemed to have helped a lot, thanks StephenB!. The connection is more stable, but still is disconnected every so often. I'm seeing a few series of disconnections last night and some now this morning. But at least it is coming back by itself.
Major improvement, but still something is not 100% ok. But at least this would be workable.
Does anyone know what could be causing this and how to resolve this?
If the ReadyNAS MTU is set to 1500, try lowering it to 1492.
That seemed to have helped a lot, thanks StephenB!. The connection is more stable, but still is disconnected every so often. I'm seeing a few series of disconnections last night and some now this morning. But at least it is coming back by itself.
Major improvement, but still something is not 100% ok. But at least this would be workable.
1492 is the usual value for PPPoE, but it might need to be a bit lower still.
You can determine the value using ping - in a Windows PC you run CMD (command prompt) and then enter
ping -l 1464 -f 8.8.8.8
If you see a message that says
Packet needs to be fragmented but DF set.
then lower the 1464 value until you find the largest value that doesn't give that error. Then add 28 to the final value, and set the MTU to that size.
Another possibility is that something else on your network might be using the default 1500. It is possible to set the MTU on PCs (both MacOS and Windows). But unfortunately I don't think you can manually set the MTU for Android (without rooting the phone) or iOS. The MTU can be signaled via DHCP, so if your devices are using DHCP it might not need to be manually set.
Does anyone know what could be causing this and how to resolve this?
If the ReadyNAS MTU is set to 1500, try lowering it to 1492.
That seemed to have helped a lot, thanks StephenB!. The connection is more stable, but still is disconnected every so often. I'm seeing a few series of disconnections last night and some now this morning. But at least it is coming back by itself.
Major improvement, but still something is not 100% ok. But at least this would be workable.
1492 is the usual value for PPPoE, but it might need to be a bit lower still.
You can determine the value using ping - in a Windows PC you run CMD (command prompt) and then enter
ping -l 1464 -f 8.8.8.8
If you see a message that says
Packet needs to be fragmented but DF set.
then lower the 1464 value until you find the largest value that doesn't give that error. Then add 28 to the final value, and set the MTU to that size.
Another possibility is that something else on your network might be using the default 1500. It is possible to set the MTU on PCs (both MacOS and Windows). But unfortunately I don't think you can manually set the MTU for Android (without rooting the phone) or iOS. The MTU can be signaled via DHCP, so if your devices are using DHCP it might not need to be manually set.
When I ping with 1464, it comes through in one piece. No error messages. So it seems that MTU set to 1492 should work.
And I spoke to soon as well, it did work for a while, even though there were a few short interuptions, but after a full day it went back to it's old state, meaning constant disconnects.
The moment I physically disconnect the NAS, they stop and the connection goes stable again. Or if I disallow the NAS access to the WAN, it also remains stable. And this is 100% reproducable. Which means it has to be the Readynas somehow.
Is it possible the ReadyNas is trying to make it's own PPPoE connection? That was one suggestion I got from AVM, but I do not know how to check this.
The moment I physically disconnect the NAS, they stop and the connection goes stable again. Or if I disallow the NAS access to the WAN, it also remains stable. And this is 100% reproducable. Which means it has to be the Readynas somehow.
Is it possible the ReadyNas is trying to make it's own PPPoE connection? That was one suggestion I got from AVM, but I do not know how to check this.
No, it won't be making it's own PPPoE connection.
What apps and services do you have running on the NAS?
Are you forwarding ports to it (or placing it in the DMZ)?
Are you using DHCP or is it manually configured to use a static iP address?
Are you using both ethernet connections, or just one?
The moment I physically disconnect the NAS, they stop and the connection goes stable again. Or if I disallow the NAS access to the WAN, it also remains stable. And this is 100% reproducable. Which means it has to be the Readynas somehow.
Is it possible the ReadyNas is trying to make it's own PPPoE connection? That was one suggestion I got from AVM, but I do not know how to check this.
No, it won't be making it's own PPPoE connection.
What apps and services do you have running on the NAS?
I have disabled most Apps. Only one still running is the Plex server App. I can test with all Apps disabled later today.
Services I will double check, but for sure SMB, FTP, NFS, ReadyDLNA, UPnP and SSH.
No cloud services in use.
Are you forwarding ports to it (or placing it in the DMZ)?
I have not set this up with the new router yet, though I have briefly tried the http port when the network seemed stable and that did work, but I disabled it again when I disconnected the NAS from the WAN.
Are you using DHCP or is it manually configured to use a static iP address?
it is being assigned by DHCP (AVM Fritzbox is the router and gateway)
and I can still access the NAS while it's connection to the WAN is cut off on the router
Are you using both ethernet connections, or just one?
I am using only one of the connections (but have tried with only the second connection instead of only the first)
Re: Readynas interrupting internet connection on VDSL
I disabled IPv6 on the NAS, as well as on the modem. In parallel I was discussing this issue with the AVM support and they mentioned to disbale the IPv6 as that seemed to not be offered by my ISP. I used their installation code to install the correct settings and then updated the firmware, which according to them must have enabled the IPv6.
Having done this, as well as switching off UpnP and the Plex app the connection seems stable so far. I will check later tonight if there have not been any disconnects.
The question then is, is completely disabling the IPv6 enough and can I switch back on UPnP and Plex. That will be the next step of the investigation.
Re: Readynas interrupting internet connection on VDSL
Clearly I spoke to soon. After a day with no issues, the connection was interupted again. Reconnecting and disconnecting every minute or so, same issue as before.
Stopping the NAS from accessing the WAN in the modem settings, immediately fixed it as before. Giving it access again after, starts the disconnection cycles once more.
I'm back at the start again unfortunately. Is there a way to see what kind of traffic there is between the modem and the NAS? And what might it want to do to break up the internet connection.
Just to clarify, the DSL line connection is not actually going down, it is only the internet connection that goes down.
I'm back at the start again unfortunately. Is there a way to see what kind of traffic there is between the modem and the NAS? And what might it want to do to break up the internet connection.
There is an audit facility you could try (you'll find it on the system->logs page, as well as system->settings->services). I don't think that will tell you much, but you should see the IP addresses of the destination.
With apps and upnp disabled, all there should be is NTP (time sync), and the NAS connecting to Netgear occasionally to look for hotfixes and availablity of firmware updates. You can disable NTP (or change the time servers), but you can't disable the hotfix checks.
Have you engaged your ISP?
Is the NAS the only ethernet-connected device on your network?
Re: Readynas interrupting internet connection on VDSL
I'll disable NTP and try to see what IP it is trying to reach using the audit facility.
I have tried to contact my ISP when this happened initially, their reaction was: hmm, I can't tell what's wrong, so it is probably something wrong with the DSL line. Let me make you a technician appointment.
Then I did some investigations which let me to conclude a connected powerline device was causing issues. At least the modem was clear in the error messages, stating that powerline devices can cause these issues. So I disconnected those and the issues went away, after which the ISP concluded the issue with the DSL line 'went away' and they canceled the technician. Which in my eyes was already unnecessary as I was 99.9% sure it was not an issue with the DSL line. So I am really not expecting any solid technical advice from my ISP unfortunately.
I am still talking with the modem provider (AVM) and they seem quite helpful. This weekend I will send them some logs of the modem in the error state (along with a full network map), perhaps they can give me some ideas what is going on.
The NAS is not the only ethernet connected device. From the modem I have a line going to the livingroom, where a switch is connected to a smart TV, a HUE hub and a sattelite box. And I have several WIFI devices connected (smart plugs and phones obviously). A second line is going from the modem to a second switch, where I have the NAS and a PC connected to. The PC is not permanently switched on.
I can try to replicate the disconnects with everything (but my laptop - need something to check and look at the error logs) disconnected.
One thing to note, is that between the modem and the second switch (connected to the NAS) is quite some length of cabling. The modem recognizes this only as a 100Mb connection, where it sees the connection to the living room switch as a gigabit connection.
I have yet to figure out, if it is indeed the length of the wiring, or perhaps wrong cabling used (though in that case also the other LAN connection should be 100Mb). Would it make sense to try the NAS on a gigabit connection, i.e. move it closer to the modem?
I have yet to figure out, if it is indeed the length of the wiring, or perhaps wrong cabling used (though in that case also the other LAN connection should be 100Mb). Would it make sense to try the NAS on a gigabit connection, i.e. move it closer to the modem?
The distance spec is 100m - likely longer than your cable run.
You do want at least CAT5e cable - if the cable is CAT5 you'd end up with 100 megabit. A broken connection in the wire (or connector) could also end up negotiating fast ethernet. You could also try a different switch port on both the AVM and the switch for that connection.
I think it does make sense to try moving the NAS - ideally connecting it directly to an AVM LAN port.
On NTP - I'm assuming you are based in Europe. If so, you could try reconfiguring the NTP servers used in the NAS to 0.europe.pool.ntp.org and 1.europe.pool.ntp.org. That does no harm, and would give you lower latency for NTP than using the Netgear servers (making the time slightly more accurate). That's configured using the device time settings wheel on the system->overview page. There are other options here: https://www.pool.ntp.org/zone/europe
I believe the NAS update servers are hosted in the Amazon cloud.
One possibility is that your NAS was hacked somewhere along the way. If you want to test that possibility, you can remove the disks from your NAS (labeling by slot), and do a fresh factory install on a single disk. Whatever is on that disk will be deleted, so ideally it would be a scratch disk (and it shouldn't be part of your current array). It's just temporary, so there is no need to worry about the hardware compatibility list. If the problem disappears on a clean install, then you should back up the files, do a factory default, and set up the system again.
With apps and upnp disabled, all there should be is NTP (time sync), and the NAS connecting to Netgear occasionally to look for hotfixes and availablity of firmware updates. You can disable NTP (or change the time servers), but you can't disable the hotfix checks.
Both of these services would use DNS of course. If you are using DHCP, then the DNS server would be using whatever the ISP is providing the AVM. I think it's unlikely to be related to DNS, but you could reconfigure the NAS to use Cloudflare (1.1.1.1) or Google (8.8.8.8) open DNS servers if you want to rule that out.
If the NAS is using a static address, then you should check the DNS server settings anyway (you don't want to be using the servers from your old ISP).
I have yet to figure out, if it is indeed the length of the wiring, or perhaps wrong cabling used (though in that case also the other LAN connection should be 100Mb). Would it make sense to try the NAS on a gigabit connection, i.e. move it closer to the modem?
The distance spec is 100m - likely longer than your cable run.
You do want at least CAT5e cable - if the cable is CAT5 you'd end up with 100 megabit. A broken connection in the wire (or connector) could also end up negotiating fast ethernet. You could also try a different switch port on both the AVM and the switch for that connection.
I think it does make sense to try moving the NAS - ideally connecting it directly to an AVM LAN port.
I know it is Cat5e cable, it is the same cable running to the attic as it is to the living room and indeed should not run over 100m. I am pretty sure, the cables after this (to the switch and the NAS) are Cat5e as well, but will double check. I have tried a different port on the AVM (the same one that is telling me it is 1Gb to the living room) with the same Fast ethernet connection. So it must be the cabling or the switch. I'll do some testing on this later today.
On NTP - I'm assuming you are based in Europe. If so, you could try reconfiguring the NTP servers used in the NAS to 0.europe.pool.ntp.org and 1.europe.pool.ntp.org. That does no harm, and would give you lower latency for NTP than using the Netgear servers (making the time slightly more accurate). That's configured using the device time settings wheel on the system->overview page. There are other options here: https://www.pool.ntp.org/zone/europe
I believe the NAS update servers are hosted in the Amazon cloud.
Yeah, that makes sense at least. And indeed, Europe based. Will switch to these servers.
One possibility is that your NAS was hacked somewhere along the way. If you want to test that possibility, you can remove the disks from your NAS (labeling by slot), and do a fresh factory install on a single disk. Whatever is on that disk will be deleted, so ideally it would be a scratch disk (and it shouldn't be part of your current array). It's just temporary, so there is no need to worry about the hardware compatibility list. If the problem disappears on a clean install, then you should back up the files, do a factory default, and set up the system again.
Hmmm, that would be worrying if that were the case. But I've got some spare disks laying around that I can test this theory with, should be easy enough.
I know it is Cat5e cable, it is the same cable running to the attic as it is to the living room and indeed should not run over 100m. I am pretty sure, the cables after this (to the switch and the NAS) are Cat5e as well, but will double check. I have tried a different port on the AVM (the same one that is telling me it is 1Gb to the living room) with the same Fast ethernet connection. So it must be the cabling or the switch.
To expand on this a bit - fast ethernet only uses 4 wire pairs in the cable, gigabit uses all 8. So if the cable or a connector is damaged, the gigabit autonegotiation often fails, and you end up using fast ethernet.
I know it is Cat5e cable, it is the same cable running to the attic as it is to the living room and indeed should not run over 100m. I am pretty sure, the cables after this (to the switch and the NAS) are Cat5e as well, but will double check. I have tried a different port on the AVM (the same one that is telling me it is 1Gb to the living room) with the same Fast ethernet connection. So it must be the cabling or the switch.
To expand on this a bit - fast ethernet only uses 4 wire pairs in the cable, gigabit uses all 8. So if the cable or a connector is damaged, the gigabit autonegotiation often fails, and you end up using fast ethernet.
Thanks! I'll have a look at this as well while I am at and figure out where this problem is located.
Both of these services would use DNS of course. If you are using DHCP, then the DNS server would be using whatever the ISP is providing the AVM. I think it's unlikely to be related to DNS, but you could reconfigure the NAS to use Cloudflare (1.1.1.1) or Google (8.8.8.8) open DNS servers if you want to rule that out.
If the NAS is using a static address, then you should check the DNS server settings anyway (you don't want to be using the servers from your old ISP).
It is set-up to use DHCP, but making sure it is assigned the same IP by the modem. So this should not be an issue and as far as I could tell, the DNS servers were reassigned indeed.
