Slower FTP file moves over a home network
Slower FTP file moves over a home network
Yes, I wanted my main PC to act as an FTP server. As you mentioned, it might cause issues if the router or firewall is blocking things. So, I added IPERF3 into the firewall and ran another test using both my mobile phone (which connects via 5Ghz WiFi) and my main PC (connected directly to the router with a LAN cable). Here are the results from that test: IPERF3 Test on Mobile & PC. What do you think? Is it working now? Another result came from Test 2 and 3, where the main PC was set as the host and the mobile phone was the client.
That router is too slow for its job as a home network drive (NAS). The small netbuilder tested it and found that writing data takes 10MB/s, while reading comes in at 14MB/s. You might need to change your MTU setting or format the drive using NTFS to get better speeds. It all depends on which model of Archer C5 board you have, since some versions use weaker processors. Instead of a router, you should buy a proper NAS, or a Raspberry Pi, or even an older computer to use as one.
It's weird because it works both ways for mobile but not for ethernet. Basically, you're capped at the router's speed of 100 mbps. With some extra overhead from internet connections or network issues, you might only get around 95 mbps on a port that says 100 mbps. The differences in testing are just due to how different types of wifi work. This means your problems with FTP could be caused by the software or the disk system. I haven't used FTP much recently except when working with old unix machines. For Windows file shares, you should really hit that full 100 mbps limit if you're transferring big files. Using a router as a NAS usually doesn't perform very well.
One thing to watch out for with TP-Link routers is that the same name can mean totally different things inside the box. Sometimes stores don't know what's actually inside either, and sometimes several versions come at the same price. For example, a model called TL-ER6120 used to sell in three different versions for the same cost. One had more RAM and a completely new processor, while others had less memory and older chips. This means you might buy a router with 360 Mbps, another with 640 Mbps, or even 980 Mbps NAT speed for just 250 to 300 dollars. You could be very glad or really upset depending on which version you end up getting. Even though TP-Link lists the hardware revision in small letters on the barcode and explains it in their specs, they rarely tell customers if retailers get those details right. Back when I was buying one for a friend years ago to set up his business network on a tight budget, I had to visit four shops before finding one willing to check what hardware version the units were actually sold with.
I hope tplink sells junk so I can tell everyone not to buy it. Their engineers need to get a big stick and visit marketing. They did even worse by using old AV500 parts under a different name like pa4010 when new AV2-600 came out they changed the name to av600 but kept calling it pa4010 so people are confused. You still find both av500 and av600 units with the same part number. These look exactly like the same thing.
I have a TP-Link Archer C1200 V2 router. I looked at the company's website, but there is very little information about the inside parts of it. They listed three different versions named V1, V2, and V3, but they are only in the support page. On the main home page of the router, they didn't mention these differences. This means all three have exactly the same hardware setup: one single-core processor, two Gigabit Ethernet ports for internet (WAN) and four for your local devices (LAN), plus a USB port. It supports certain file systems like NTFS or FAT32, lets you run FTP servers, use Samba, has WPS to connect quickly, and includes power switches. The router uses 12 volts from the wall.
So it seems like this router only has 100m ports. Maybe that's why you said the speed was 100/100. If so, you probably have broken cables or they are set to something other than auto mode. Either way, this will make your iperf tests go really slow. You should get more than 900mbps with gigabit ports and iperf. I don't know if that matters for your FTP speeds since you're still well under 100mbps. Maybe the errors are from a bad cable? The vendors don't explain much about what makes versions different. There are other databases where people take apart devices to read reports on the FCC site, which gives way more info. The main difference between those three models is that they use different CPU brands and chips. This mostly matters for third-party firmware like dd-wrt because some chip brands don't support any modifications. In your case, the big difference is things like clock speed and memory size. The CPU makes a huge impact on how fast FTP servers run. Unfortunately, you can't directly compare clock speeds between different brands of chips. From what I can tell, v2 is actually the strongest version. The later models used a chip with a much slower clock and cut 75% of the memory by making it smaller.
You were right—the USB port on the router isn't really very good; it feels more like a fake extra to show off that things are better than they are. A real dedicated NAS is what you need instead. I built one using TrueNAS Core on an old PC that had been sitting in my dad's office all unused. Now, every device on the network gets full speed at 1000/1000 Mb/s. Building it up and learning about how a NAS works has been super rewarding and fun to explore other things like running virtual machines or setting up my own cloud storage.