@Carl 

As I was the one that initially recommended turning off Mesh Wifi for those affected I am glad to see it has worked for some. Improving the performance of the Deco devices. If you keep having issues with latency after this "fix" I would just return the device and go for a different company as it is clear TP-Link customer service does not understand how intermittent ping spikes can affect gaming. For example, if I'm playing the online game I prefer and have a spike in the 3000ms range, I just get disconnected. Doesn't even matter if it happens once in a while. It can be devastating. Especially in a competitive match. Average ping is nothing if you have spikes. What we need is consistent ping. It hurts me to see that "networking experts" at TP-Link can't understand this basic requirement and have not found or seem to desire to find a solution. I will not be recommending or buying from this company in the future due to the unprofessional and unknowledgeable service representatives and broken products.

@Carl I guess it depends on the game. I mostly play League of Legends. If I have an average ping of 100ms, it sucks, but I can deal with it. However when I get the random spikes I was getting with the Deco to 2-300ms, it costs me my life and could easily cost my team the game. Not as big of a deal in a game like COD I suppose where you just respawn instantly. Oh well, happy gaming everybody!

@Warthog710 

No one ever stated that turning off the Mesh was a fix. It helped you, that’s great, but we also need to collect more information, as was stated in my response to @ausinuk yesterday.  We aren't able to address a concern without having full context of the overall situation.  For every network action there are at least 5 points of reference that are involved and can affect an applications performance.  Those being the service, the modem, the router, the client device, and points of connection.  Also, you have to consider load, demand, range, interference, congestion, among other things. If we just look at one aspect, ping times, without knowing the whole picture we can't troubleshoot or determine a solution.  The truth of the matter is that spikes in your ping times are going happen, regardless of the router you are using, the game you’re playing or application you have active.  This is especially true if you are using wireless to game. Building off that we also have to look at if it’s just one client device having issues or multiple.  If it’s just one device its typically the network adapter on the client device that is the culprit.  We look at the average ping time and not the maximum, because we are looking for something that is happening consistently.  The maximum and minimum ping times don't show this.  This is also why I have customers do a loaded packet ping test of at least 500 packets, so that we get more true to life real world results.  If you can provide us more details we would be happy to look into your particular concern further.

@Carl 

I never claimed turning off mesh wifi was a fix. If you look at my post you may notice I put fix in quotes. A customer should not be forced to turn off a main feature of the device to fix a problem that shouldn't be occurring. As to your belief that maximum ping isn't as concerning as average ping. Let me give you some numbers from one of my previous tests. I sent 50 packets to google's servers. Each of these packets contained 256bytes of information. I averaged 114ms. According to your previous posts, this is quite good and acceptable. However, my maximum ping was 3614ms and 3 of the packets were never returned! This is a case where the maximum ping is extremely concerning. If I was to go to a customer with a product and told them that 95% of the time it will perform well. But the other 5% it fails. Depending on the application this is completely unacceptable. Networking devices (especially in the modern world) need to be extremely reliable as many people rely on them for day to day activity. Having something that commonly fails or produces headaches for the customer is not good. Especially in a market crowded with competitors. As to your claim that spikes are just a terrible truth for wireless networking, why does this not occur on other routers from other companies? I realize this is a terribly difficult thing to troubleshoot. Anything that only occurs sometimes is extremely hard to figure out. Believe me, I know this personally in my work. But the cold hard truth is that these problems need to be solved still. It's not fair to the customer to release a product that clearly fails to produce comparable results to other companies' offerings and claim it to be just as good and/or better. Browsing your forums, it seems that this issue is fairly common. This should be a high priority issue for your engineers to replicate and solve. 

@Warthog710 

Your right it is incredibly difficult to test concerns like this.  This is why we ask for as much detail as possible.  So, we can start to narrow down and isolate the concern and find the cause. Just because there is a spike in the ping or occasional packet loss does not mean the router is the cause of the concern.  It might be an issue with the service, the ISP hardware, the client device or its adapters or even the cables connecting devices.  Usually when having issues of high pings and/or packet loss we ask to be contacted when the issue is happening because that is when its most reasonable to find the cause.    For instance, you could be seeing high ping spike when playing a MOBA but at the same time another person on your network is streaming a 4k movie. You could even be on an ISP who does shared bandwidth like cable providers and there is an increase in usage causing bandwidth strains in your area (this is my current situation).  You could have an undiagnosed line issue or a short in an ethernet cable. There could be congestion or interference with your wireless affecting the signal of the wireless service you connect to.  This is very common for users who use a closed enclosure for their PC tower.  The point is there are several things to look at to find a cause. While truthfully an occasionally high ping is not much concern, consistently high ping spikes can be and are something that we are ready and fully capable of helping to address, but to do so we need as much information possible.  

Right now, we are just going back and forth and not really addressing the issue at the heart of your concern.  I've explained our standpoint and you have explained yours. For you or anyone else that needs or wants to address this particular concern please run the following tests.  Run these tests the following ways A. Wired and Wireless from the base node.  B. Wired and Wireless from a remote node C. Repeat A and B on a separate computer

• Ping <Router's gateway> -l 500 n 500
• Ping <public IP address> -500 n 500 *make sure the site allows 500 byte packets
• Tracert <public Ip> if you can trace to your primary game's server that would be even better.

Also, we need to confirm 

• how often this happens. 
• If its consistent
• If it’s at a specific time of day 
• How many devices are on the network
• How many of these devices are doing bandwidth heavy tasks

That should be enough to start with.  Please respond with the results and we can go from there. 

@Ashhong, I also have the same issue.

Model: Deco E4R

Hardware Version:

Firmware Version: 1.1.1 Build 20191216 Rel. 62031

 

Client Adapter: TP-Link Archer T2U AC600 Nano and Plus

 

 

Please help!

@Carl Here is the information you required.

Operating mode  - Access Point

Current firmware: 1.4.4 Build 20200221 Rel 65392

Deco M5 of which there are 3 in the house

Ping results with Ping <gateway> -l 500 " mesh turned off"

Ping results with Ping <gateway> -l 500 " mesh turned on"

Ping results with Ping www.google.com -l 500 " mesh turned off"

Ping results with Ping <gateway> -l 500 " mesh turned on"

Tracert to www.google.com mess off

Tracert to www.google.com mess on

So to your questions

• how often this happens. Every single time mesh is turned on for any device connected to Deco WiFi. With Mesh off very inconsistent
• If its consistent - very consistent and it does not matter if this device or any other device is connected to any of the Deco M5 devices. Same situation. With Mesh off very inconsistent
• If it’s at a specific time of day - No anytime of day or night
• How many devices are on the network - 18 devices across 3 Deco M5
• How many of these devices are doing bandwidth heavy tasks - during the taken of these results I was the only one on the network using a computer. 

I will point out as previously mentioned. If I point any of the 18 devices to my existing Tp-Link router AC3200 I do not get any problem what so ever. Ok so why not just use the Tp-link AC3200 router? The reason behind the decision to buy a mesh network was to get wifi up into my loft (my office and gaming area). Obviously the AC3200 does not do mesh. I actually decide to go Deco, against colleague recommendations, because Deco is tp-link so I thought it would play nicely with the AC3200 etc. I like to stick to similar brands where possible in the hope that they do play nicely together. 

I hope this helps and please turing "mesh off"is not a solution. It is a band aid to a problem in the mesh capability. As I also mentioned that turning off mesh does not always gaurantee me no spikes but they are certainly more random. Hence you will not see them in the results above.

Regards

Ausinuk

@Edoe 

See the steps i gave in my response above yours.   Please run these same test and provide the results. 

@ausinuk 

Yes we agree turning mesh off is not a solution. 

A few more questions:

• Since you are using the Deco in tandem with the C2300 is the Deco in AP or Router mode?
• When the testing was performed were you connected to the Base node or a remote one?
• Can you run the same test for the opposite as well.  This way we have a better idea of the whole picture.  
• Where the test ran while connected via ethernet or wireless?
• as with testing from the node can you also if possible test the opposite way as well.  Again it gives us a full picture.
• For the ping to a public IP address could you run it again but to something other than Google?  The reason is if you look at the results google reduces the packet size to 64 bytes instead of 500.  Just as long as the site allows 500 bytes it will be good 
• Finally can you post the results of the pings to the gateway and public.

thank you 

@Carl 

• Since you are using the Deco in tandem with the C2300 is the Deco in AP or Router mode? Dec is in AP mode
• When the testing was performed were you connected to the Base node or a remote one? Remote node
• Can you run the same test for the opposite as well.  This way we have a better idea of the whole picture.  Below re the results from wifi connection to main Deco which is connected directly to Ac3200 router.
• Ping gateway (main deco router) mesh on
• 

Ping gateway (main deco router) mesh off

• Where the test ran while connected via ethernet or wireless? all results wireless
• as with testing from the node can you also if possible test the opposite way as well.  Again it gives us a full picture. not sure what you mean?
• For the ping to a public IP address could you run it again but to something other than Google?  The reason is if you look at the results google reduces the packet size to 64 bytes instead of 500.  Just as long as the site allows 500 bytes it will be good 
• Ping to www.oracle.com mesh on
• 
• Ping to www.oracle.com mesh off
• 
• Finally can you post the results of the pings to the gateway and public. I assume the ones in this message is what you meant.