Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[bjn]

I recently added a DAP-1522 to my homem network (DIR-655 router) in bridge mode.  I have found using LanSpeed that the read speed is much slower than the write speed.  The write speeds are 21 Mbps and the read is only 8 Mbps.  I thought this was because of some Windows issue, but then I set up a Linux server and I see the same thing.  When I take the same computer and disconnect from the DAP-1522 and turn on wireless I get read/write of 21 Mbps. 

1. I know it is not a poor connection as I have attempted the tests in the same room and confirmed strong signals.
2. It is not an issue with the servers since I can use wireless on the same computer that was connected to the DAP-1522 and the issue goes away. 
3. The DAP-1522 is in mixed N/G mode, but I have already tried just using N and I see the SAME issue.   
4.  I have tried changing channels, set the DAP-1522/DIR-655 to WAP2/AES
5.  I upgraded to the latest firmware for both the DAP-1522 and DIR-655

Anyone have any ideas on what could be causing it?  I am really disappointed in the DAP-1522.

[tamsler]

I have two DAP-1522 (one as an AP an the other one as a Bridge) and did some speed tests. Setup:

- 1 wired linux system
- 1 wired or wifi macbook pro
- AP accepts only N, and connects to Bridge via 5GHz and
- CAT 5 ethernet cables, which should be good for at least 12.5MB/s (100Mb/s)

Test 1 AP to wire:
scp 100M file: highest throughput was 1.0MB/s (8Mb/s)

Test 2 Bridge to wire:
scp 100M file: highest throughput was 1.0MB/s (8Mb/s)

Test 3 wire to wire on AP:
scp 100M file: highest throughput was 1.1MB/s (8.8Mb/s)

The result indicate that the transfer rate is just a little bellow 1.25Mb/s (10Mb/s). I would have expected that "Test 3 (wire to wire)" would have resulted in a higher transfer rate, closter to 12.5 MB/s (100Mb/s)

Best,
-- Thomas

[bjn]

Thomas,

Thanks for the data, although I am not sure its quite the same as what I was doing.  Might just be I am misreading what you did.  Can you try this for me:  take a file from the MAC wired to the bridge and write it to the wired linux system (I assume the linux is connected to your router and not the DAP-1522).  Then, read the same file from the linux system back to the MAC wired to the bridge.  That would be equivalent to what I did.  Now, if you did the same experiment a second time, but the MAC could connect wirelessly to the router without being connected to the bridge and you would duplicate my results.  i.e.

ComputerA wired to bridge that is wirelessly connected to DIR-655 => writing to computerB wired to DIR-655   22 Mbps
ComputerA wired to bridge that is wirelessly connected to DIR-655 => reading from computerB  wired to DIR-655   8 Mbps
ComputerA wirelessly connected to DIR-655 => writing to computerB wired to DIR-655   22 Mbps
ComputerA wirelessly connected to DIR-655 => reading from computerB wired to DIR-655   22 Mbps

Hopefully someone will have an idea on why there is a difference

[tamsler]

I tried to reproduce your results and mine a a little bit different:

Note: I have two DAP-1522 in my setup, one is an AP and the second one is a bridge

Test Case 1:
Mac wired to DAP-1522 Bridge switch
Linux wired to DAP-1522 AP switch
write 100M file from Mac to Linux: 25.6Mb/s

Test Case 2:
Mac wired to DAP-1522 Bridge switch
Linux wired to DAP-1522 AP switch
write 100M file from Linux to Mac: 8Mb/s

Test Case 3:
Mac Wifi (N) to DAP-1522 AP
Linux wired to DAP-1522 AP switch
write 100M file from Mac to Linux: 31.2Mb/s

Test Case 4:
Mac Wifi (N) to DAP-1522 AP
Linux wired to DAP-1522 AP switch
write 100M file from Linux to Mac: 8Mb/s

I think our test results are the same except for my Test Case 4. I am not quite sure why there is a difference. In my case, My MAC is quite a bit more powerful then my Linux system in terms of CPU, disk, etc. So I don't know if that's the issue.

Best,
-- Thomas

[bjn]

Thanks Thomas,  that confirms what I am seeing.  When writing from CompA to CompB through the DAP-1522 the speed is relatively high, but when reading it drops down significantly.  I think if you had a wireless router instead of the DAP-1522 as an AP then you would see similar to what I saw in your test4 as your first two tests confirm my issue.  Something is not right with the DAP-1522 and reads.  Anyone have any idea why?

[jerry6]

try the 5ghz band if you can get a strong signal , low throughput on 2.4 is a fault 
http://www.smallnetbuilder.com/wireless/wireless-reviews/30455-d-link-dap-1522-review-dual-band-draft-11n-for-the-masses?showall=&start=5

[bjn]

Jerry6,

The DIR-655 does not have 5ghz, but thanks for the information.   In general the writes are what I would expect from the link you provided, 21 Mbps.  Would you expect the writing of a file going through the bridge and the reading of the same file to have different throughput from the test you did? 

As far as I could tell from the link you provided it was just reporting poor performance when going to 2.4ghz, but I am getting this "optimal" performance, but only during file writes.  My issue is that I get poor performance when transfering files in one direction, but not the other direction.  If I write information and it goes through the bridge it is ok (even if it is not as fast as I would like), but if I read a file going through the bridge it is 3X slower.     

try the 5ghz band if you can get a strong signal , low throughput on 2.4 is a fault 
http://www.smallnetbuilder.com/wireless/wireless-reviews/30455-d-link-dap-1522-review-dual-band-draft-11n-for-the-masses?showall=&start=5

[tamsler]

I found the solution for the slow throughput. I had the netmask misconfigured on my linux box. I had 255.255.255.255 instead of 255.255.255.0. After fixing that, everything works perfectly.

Best,
-- Thomas

[bjn]

I'll check that but pretty sure it is not my issue.  I tried another experiment in which I hooked the DAP-1522 up to a Belkin Powerline adapter and used it as an AP.  Same issue.  I also tried a different set of computers.  This is really frustrating.  I was seeing something similar when I used a DIR-628 in place of the DAP-1522 (in wired mode) and fixed it by turning off all the file optimization methods.  I can not seem to do this with the DAP-1522.