Posted: Mon Jun 17, 2019 9:29 Post subject: WAN packets drop WRT3200ACM [SOLVED]
I'm the unhappy owner of a WRT3200ACM.
Let me explain my quick story.
Several years ago, I bought a Asus RT-AC66U which was supposed to be good but at the end, it was the first release and has 1 core and after a while, it become laggy and ping to 1s. There is no more support at DD-WRT for that router and I finally flashed a ... fork of a fork and it's not better, still laggy after a while. But no packet drop (and I'm testing it again now as the Linksys is just horrible).
Note that I also tested reboot each day but it can fail in less than a day so ...
So I bought a brand new WRT3200ACM on Amazon and upgraded to latest Linksys and just used it with stock firmware. I discovered that there were connection dropped and so, I started to play with ping on machine connected to wire and Wifi and it appears that Wifi and wire are OK to connect from LAN to LAN but once you have to go through the WAN port (to connect to the router of my ISP), than there is a high rate of packet drop and connection seems to be stuck for 10s or more.
With the Linksys firmware, both IP of the router (LAN and WAN) were having different behavior, I mean that the LAN ip was OK and no packet drop the but WAN ip has packet drop. That's why I first suspected that the stock firmware was having issue, I also found others having the same problem and heard to flash to DD-WRT to "avoid that unstability".
Ok, I flashed DD-WRT yesterday but unfortunately, the problem persist and I still see 10s of dropping packet and so, about 3% packet loss on longer run. I also (surprisingly) found that the LAN and the WAN ip have no issue at ping (not like the Linksys stock firmware) and so, I think the problem is a bit "less worst" on DD-WRT that on the Linksys firmware. But, I still see packet drop when pinging the ISP router (so, physically out of WAN). I put back the previous router (with ping to 1s after a while) and that one has definitely no issue with packet loss and so, it's not the ISP router which is failing as it works with my old router. I also changed the cable between routers (Linksys and ISP) without more luck.
So I could still do a last test : connect an external switch to the WAN port and add another device (other than my ISP device) to ensure that it's not a conflict between those 2 devices (hell why ??).
Here is the story, at that point, I'm stuck with a defect router (or did I miss some configuration which could cause that issue ?) and from above analysis, I can only conclude that my router is hardware defective with the WAN port or something related.
I would like to add that at first install of Linksys router, it's so annoying that .. yeah, it can't live without internet access, really annoying. Then after several days, it started to be slower, I think it rejected Wifi clients and I was not able to connect to the Web interface. I restarted it and .. nothing, it lost my configuration ! *shocked* I had to reload a backup (fortunately I've took a backup at the end of my "basic" configuration). Yes, I'm using very few customized settings and it lost it. After loading back config, it was working correctly, excepted packets drop on WAN.
I've no idea of what's happening, the Amazon return window is done and I will need to go for warranty and I'm super disapointed to see that a 200 euro router is no more able to simply route some ping packet without trouble. My old Asus and Linksys combined, it's been years of network issue.
I would like to get some feedback about that, do you think I had a chance to get a defective one or more configuration issue ? Also, I think my router is the last version (the one with locked radio config) but I'm not sure and there is only 1 DD-WRT firmware for wrt3200acm.
Thanks u for your input
Last edited by loop_x on Sun Jun 23, 2019 20:56; edited 1 time in total
Joined: 17 Feb 2010 Posts: 599 Location: Yorkshire (GOC)
Posted: Mon Jun 17, 2019 9:44 Post subject:
If you are new to DD-WRT, then it is almost certainly your configuration There are plenty of WRT3200ACM users here and reports of serious issues are rare with this model. I have a WRT32X which is only a slightly different spec. and it has been rock-solid for me.
Golden rule for DD-WRT... NEVER slect "Auto" for anything important!
There are loads of posts in this forum about WiFi and Security configurations, so I won't bore you with those.
The obvious key question is do you see any WAN packet issues over a purely wired connection? If so, then it may be your ISP connection that is flakey. This would be the first thing to resolve.
BTW, r40009 seems to be a good solid DD-WRT build to start with. _________________ Linksys WRT32X v1 - r43217
Linksys WRT1900AC v1 - r43217
TP-Link Archer C9 v1 - r43217
I made test and the Wifi works perfectlly, it's just the WAN output which is failing randomly.
Also, it was the same behavior NOT with DD-WRT so, not sure it could help to set some setting to non-auto with DD-WRT.
Also, I'm using the 40009 revision : first install, no issue except the drop packet.
Do you have any idea of what setting could cause such issue with the WAN ? Again, remember it' not a Wifi or Wired issue, all is OK, just going through the gateway (WAN) is failing and my old router is not failing so, it' not the ISP router which should be an issue. MTU should also be set to 1500 as usual, no idea of what option to try
EDIT: just tested with MTU at 1000 on WAN and obviously it's not working better and I see packet ... errors now and no more on br0 but on eth0.
So I just restarted (after MTU to 1000) and it's failing already after 4 min uptime. Do you really think it could not be a hardware issue ? Having that problem with stock and ddwrt, seeing drop on br0, error on eth0, this router has just never worked.
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MULTICAST MTU:65536 Metric:1
RX packets:4 errors:0 dropped:0 overruns:0 frame:0
TX packets:4 errors:0 dropped:0 overruns:0 carrier:0
RX bytes:757 (757.0 B) TX bytes:757 (757.0 B)
EDIT2 : tested with MTU WAN at 1462 (as tested with ping stuff) and now, no more error, but back to dropped packed on br0 .. can someone understand that ?
EDIT3: tested a ping directly from the router, what I see :
- also impacted but seems to lose 3 packets less than the wired PC
- the PC throw destination host unreachable
- the router did no error, just packet loss and once back, first packed has 2s ping and second was 1 then back to normal (same on PC but a third packet with 45ms ...)
EDIT4: no in fact, about the same on PC and router once it fails and each time I see a failure, it's about 10 packets drop (so about 10s) and on all devices are impacted at the same time, then back to normal. This really makes Skype crazy :-/
EDIT5 : I contacted (chat) Amazon DE and 10 days after closure for returning packet, they still have me labels for return I'm glad, I'm gonna push that router away from me and wait a new one which should fix the issue as I don't see what could be the problem.
On the Linksys stock and updated firmware : looks like the problem persist. Exactly as the previous one, I have packet loss ONLY starting at WAN or beyond BUT no problem to ping the LAN interface of that router. So the routing is not working well and is the cause of those drops.
I will try DD-WRT but it's already not looking good.
Also here are some results :
- wired PC to LAN IP of the router :
2714 packets transmitted, 2714 received, 0% packet loss, time 2713305ms
=> cable works fine
=> if Wifi, it's the same
- wired PC to WAN IP of the router :
2743 packets transmitted, 2710 received, +1 errors, 1% packet loss, time 2742349ms
=> here is the problem
=> and it's the same device ... but that IP can't ping without loss (with DD-WRT, it will work though)
=> also I remember that with the stock firmware, there are "errors" where there are no "errors" in DD-WRT (just packet loss)
- wired PC to private IP of my ISP (so, just behind the WAN) :
2720 packets transmitted, 2680 received, 1% packet loss, time 2719417ms
- wired PC to 184.108.40.206
2724 packets transmitted, 2676 received, +2 errors, 1% packet loss, time 2727163ms
It seems that the packet drop could be a bit less than before, but I need to confirm with DD-WRT.
Globally, the problem seems NOT to be fixed at all, so I will probably return also that router and never buy anything else from Linksys as I'm done with routing issue with a 200 euro router.
Can you post the configuration of your WAN setup? Under advanced-routing, is your router in gateway mode? I don't lose any packets pinging my WAN interface from my inside network, so something isn't right in your environment.
The problem is identical on both firmware : stock and DD-WRT, excepted that on the stock, problem start at IP of WAN where with DD-WRT, problem is with IP of ISP router (where the modem is ..).
PC => switch => WRT => Fritz => Internet
PC => WRT (LAN) : always OK, no packet loss
PC => WRT (WAN) : always OK, no packet loss (on stock firmware, it start failing here)
PC => Fritz : fails with dropped packets
PC => Internet (8.8.8. : same as Fritz
Once I use the previous router, no more packet loss to ping Fritz from PC.
Configuration is the follow :
- Setup / Basic Setup : Wan Connection Type : Static IP, netmask with 255.255.255.192 and Gateway + DNS set with Fritz IP
- MTU auto, Shortcut Forwarding Engine Enabled, STP disabled
- at Router IP section, I just set the IP of the LAN for WRT and not the gateway nor DNS (as already set on WAN interface)
- DHCP is disabled on the router as there is another devices managing them
- MTU set to auto, Shortcut Forwarding Engine enabled and STP disabled
- NTP set and date OK on router ...
- Advanced Routing : Gateway, Dynamic Routing (disabled), never added any static route
- the routing table display "default" with gateway to Fritz + 2 connected subnet (LAN & WLAN and the second is WAN)
- so with DD-WRT, I don't lose packet pinging the LAN/WLAN or WAN interface
- ping the Fritz is having packet loss
Fact is that I can't say that the switch is an issue nor the Fritz is an issue nor the cable between WRT and Fritz is an issue as with the previous Asus router, there is just no packet loss to ping the Fritz box and I speak about several 10k ping so, it's stable and the only thing I see is going through WRT is the issue (and same issue with stock or DD-WRT).
Also how to explain that with stock firmware, the first IP to fail is the WAN .. and every devices are impacted (cable or wifi) for several second after several minutes (but seems to be less with that new WRT but still present : from 3% I now see 1%).
0% could be seen also but it still means that there are packet loss and it means that I could encouter some connectivity issue (and yeah, with Skype when I have to work, it's an issue but have to monitor with that second new WRT).
I've no idea why it's so complicated for that router to route the packet : we speak about little "ping" packet and it fails to not lose them.
Here are more test (from wired PC) :
- to LAN ip :
1179 packets transmitted, 1179 received, 0% packet loss, time 1178032ms
- to Fritz :
1218 packets transmitted, 1213 received, 0% packet loss, time 1217056ms
- to 220.127.116.11 :
1211 packets transmitted, 1200 received, 0% packet loss, time 1211840ms
- to WAN :
1210 packets transmitted, 1210 received, 0% packet loss, time 1209060ms
So for now seems better but still packets loss.
Also the Wifi is working super fine but going out to Fritz = same packet loss ...
Crazy issue maybe power issue, should try ... moving the router away from current position ...
Will try a long ping directly from the WRT now, the problem should also be visible.
Sometimes, the ping say Host Unreachable (when pinging Fritz) but no issue with ping of LAN / WAN ... so the WRT receive the packet correctly and the firt WLAN/LAN interface are OK. But after leaving the WRT to Fritz, there is drop which was not existing on previous router.
EDIT : here are some ping result from a PC connected with WLAN :
- to LAN ip :
5242 packets transmitted, 5242 received, 0% packet loss, time 5249797ms
- to WAN ip :
5330 packets transmitted, 5330 received, 0% packet loss, time 5337821ms
- to Fritz :
5324 packets transmitted, 5283 received, +4 errors, 0% packet loss, time 5331031ms
- to 18.104.22.168 :
5410 packets transmitted, 5361 received, +3 errors, 0% packet loss, time 5417923ms
So once again as you can see, connection to WRT is perfect, but once leaving to Fritz, it's dropping. I will connect my work laptop on cable for several hours and ping it from my 2 others testing PC (LAN + WLAN), maybe the result will permit to confirm that it's not only the WRT but the Fritz box but again, it was OK with previous router (tested several days ago, not speaking about a dead device even).
EDIT2 : ok, doing another test, added 1 PC in the LAN between WRT and Fritz, added ping to that devices from my 2 PC and also running tests from inside the WRT.
So, there is 16 ping in //, let's see how it plays.
Also, as I want a good test, I added another switch after WRT (WAN) to avoid using the switch of the Fritz.
PC => Switch => WRT => Switch => Fritz | new_PC_test
About the same for Wifi :
PC_WLAN => WRT => Switch => Fritz | new_PC_test
Fritz still provide internet access.
EDIT3 :here are results :
- from PC Wired :
=> LAN IP :
10039 packets transmitted, 10039 received, 0% packet loss, time 10038265ms
rtt min/avg/max/mdev = 0.200/0.480/4.964/0.056 ms
=> WAN IP :
10077 packets transmitted, 10077 received, 0% packet loss, time 10076304ms
rtt min/avg/max/mdev = 0.214/0.480/1.073/0.031 ms
=> Fritz :
10117 packets transmitted, 10117 received, 0% packet loss, time 10116347ms
rtt min/avg/max/mdev = 0.616/0.853/1.517/0.052 ms
=> 22.214.171.124 :
10162 packets transmitted, 10161 received, 0% packet loss, time 10176315ms
rtt min/avg/max/mdev = 11.870/12.492/22.760/0.296 ms
=> test PC next to Fritz :
10209 packets transmitted, 10209 received, 0% packet loss, time 10216250ms
rtt min/avg/max/mdev = 0.640/1.132/5.349/0.260 ms
- from WRT :
=> Fritz :
10185 packets transmitted, 10185 packets received, 0% packet loss
round-trip min/avg/max = 0.317/0.420/1.923 ms
=> test PC next to Fritz :
10179 packets transmitted, 10179 packets received, 0% packet loss
round-trip min/avg/max = 0.306/0.709/1.318 ms
=> 126.96.36.199 :
10190 packets transmitted, 10190 packets received, 0% packet loss
round-trip min/avg/max = 11.542/12.063/19.944 ms
well, that's surprising !!! Since the infrastructure update by adding a switch between WRT and Fritz, it seems to be working perfectly !!! But why ???
I have been trying to figure out these problems with packets being lost while network testing. Not sure if the problem is just the nature of the beast with Nordvpn, or issues with DD-WRT OpenVPN Client.
ISP Router test: Ping very low, Jitter very low, Packet Loss 0.0% (215Mbps down, and 11.5 up. MY Normal ISP Speeds). VPN Linksys WRT3200ACM Router with DD-WRT installed, and all normal options enable including NordVPN DNS 1, DNS 2, in setup, wireless, etc, and OpenVPN Client disabled. Linksys Router test same outcome as ISP Router: Ping very low, Jitter very low, Packet Loss 0.0% (215Mbps down, and 11.5 up. MY Normal ISP Speeds). With OpenVPN Client enabled with protocol UDP test: Ping very low, Jitter very low, Packet Loss 2.6% - 36.0% 140Mbps down, and 11.0 up. TCP test: Ping high, Jitter high, Packet Loss 0.0% 35Mbps down, and 11.0 up.
Using a 100mbits switch or 1gbits switch between wrt and fritz seems to fix the problem. Discovered that the used port on the fritz is 100mbits and that 2 of them are tagged 100 and 2 others tagged 1000. On the gigabit switch, connection to fritz is 100mbits and working with that gigabit switch where it fails when I connect directly both router without switch (and yeah, it also displays 100mbits for the the connection in wrt).
I also detected that the test pc connected on the fritz switch is also failing with packet loss, and its a gigabit pc connected to 100mbits fritz, exactly like wrt. So I start to believe that the issue is not the wrt but the fritz, but could also be wrt, see below.
I start to believe that the problem could come from from those 2 possibilities :
- an issue with those 2 100mbits ports on fritz (but the old router was able to avoid issues, hard to believe then and it was a 1000 wan port also but not tested so maybe it was just a 100 and so, no issue)
- some devices are having special behaviour when enabled for gigabit and something like auto mode : I suspect a possible "retry" from the 100 to the 1000 followed by a fallback to 100 and this, periodically where some other devices are not retrying and so, have a stable connection. Those stable devices could be switches and the old router. The wrt and test pc connected also on the fritz switch want gigabit and retry periodically which break the connection.
loop_x your routers seems to be fine to me. Looks to be a DNS + fritz issue. The ISP and linksys router alone seems fine even before, and after DD-WRT is installed right? Doesn't seem to be at the moment a devices issue when testing. I only use my router for VPN so I wouldn't be much help to you.
I found the problem, finally. For now, several hours and no more packet loss .
It's absolutely not a DNS issue nor a ISP "line" issue. My test is simple ping on IP directly so, not possible DNS is the issue.
As I just write above, I found the issue. It's not because of a "retry" to get a 1Gb when connected on 100Mbits : nothing to do which that.
What I discovered is that 2 port of the Fritz box are 100Mbits AND are the problem. 2 other ports are tagged 1Gbits but 1 is really set to 1Gbits (LAN 1) and the second is 100Mbits (LAN 2). My tests clearly show that using port LAN 1 and LAN 2 are not impacted with packet drop.
The big question is why are LAN 3 and LAN 4 (the 2 who are failing on the Fritz box) working fine with previous Asus router and with 2 switchs ??? That's where I will just say that it's probably incompatible hardware for a really unknown reason : switches are OK, older devices could be fine with those Fritz port but my test PC (new) and WRT (newer) could have incompatibility issue using those LAN 3 and 4 and so need to connect to LAN 1 and 2 only (and WRT will go to LAN 1 which is also the only port available out-of-the-box with 1Gbit). The test pc will soon be disconnected as tests are now complete and I see what the h* has happened.
I sent back 1 routeur to Amazon for nothing as it was probably OK, it's really unlucky to had such issue, and firt of all, if I was aware that not all port was gigabit on that Fritz box, I would directly tried that one, LAN 1 : before that, I just connected 1 port of the switch, don't care which one as they are supposed to all be the same and was OK with previous router).
Oh well, now should be fine, thousands packets sent for tests and no one is dropped (or very very few, 3 packets over Wifi + Internet on 4k+ packets seems OK to me).
Hope no body will get that nightmare, spent hours of testing just for that, just for that ...
Enjoy your life everybody
EDIT: oh, I forgot that I moved the test PC from LAN 3 (a failing one) to LAN 2 (OK one probably on another chipset, but at 100Mbits) and it started working fine and so, nothing to do with the "refresh" things. So all match perfectly.