Print

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

[prettybird]

Earlier this week I upgraded my office's network of 5 Asus RT-AC68U's running in Access Point mode. Frankly, the old system had been working flawlessly. Whatever black magic was happening under the hood allowed roaming between AP's to be utterly seamless, rarely dropping even a single ping as I was testing. The same was essentially true for an older network of similarly configured Buffalo WZR-600DHP's.

But since I was having latency issues with our new exterior WiFi security cams, I figured it was a good excuse "upgrade" to a more legit business class solution. After some research I went ahead and bought four DAP-2695's for our almost all-Mac office in the US. One AP is set up as the master, one as the backup master, and two slaves in an AP array on the same subnet dedicated to WiFi.

- All are on the latest 1.17 firmware
- One SSID, secured with WPA Personal, for all AP's and bands
- Band Steering enabled with default values
- Channels are set to "auto" for each
- Auto-RF is enabled for the AP Array
    - Auto-Initiate Period: 24 hours
    - RSSI threshold: 40%
    - RF Report Frequency: 10 seconds
- RSSI "aging out" threshold enabled at 70%
- ACL RSSI enabled at 80%
- Beacon Interval changed to 40ms

All other settings remain default.

It has been nothing but headaches as I've tried to get these things to work, and even at their best they seem to be a huge step backward in every way beside their range and throughput, which is fantastic.

I'm hoping you can help me resolve some of these problems, or they are going right back to the seller.

1)  I can't get them to stay alive more than a day, at best. And since the logs are stored in volatile memory, I have no idea what the hell is going on when they do go down. Every connected user is kicked offline, and sometimes they can't re-associate with another AP for a long while, or at all. They're sometimes down for minutes, disconnecting network drives, kicking them off chat, and interrupting their work at random points throughout the day. I can't seem to anticipate or reproduce these reboots, and I don't know what might be causing them. They seem completely random.

2)  After moving through the office and aging out an old connection, there is often a significant delay of up to 45 seconds as the client is dropped and connects to a closer AP. Sometimes it's a lot faster, but it isn't consistent. Is there any way this can be optimized, or is this expected behavior with these units? If so, they will likely not work as a permanent solution.

2)  I don't know if this is relevant, but I cannot get the NTP settings to work or to stick. After these aggravating reboots the system time for the AP is always 1970. I've tried my office router's NTP server and a number of publicly available NTP servers, but none seem to work. Again, I'm not even sure if this causes any problems, but it doesn't say much for the stability of these units.

Thanks for taking a look at this. Any advice would be appreciated.

[FurryNutz]

Link>Welcome!

• What region are you located?
• Try setting a manual Channel to a open or unused channel. 1, 6 or 11. 11 for single mode N if the channel is clear. 13 for EU regions. Try channel 48 or 149 on 5Ghz.
• Any 2.4Ghz or 5Ghz cordless house phones or WiFi APs near by that maybe causing interferences?
• Any other WiFi routers in the area that maybe causing interferences? Link> Use a WiFi Scanner to find out. How many?
• If you have any of these options, Try turning OFF or ON Short GI, WLAN Partition, Extra Wireless Protection and HT 20/40 Co-existence if you have it. Also testing with HT20/40Mhz Co-existence enabled will impact results as well. I prefer to use this option OFF. Recommended settings are default. Under Advanced/Advanced Wireless.

What is the distance between each DAP?

Might try using just one or two DAPs and see if they continue to disconnect as a test.

Might try more recent FW updates and see:
http://forums.dlink.com/index.php?topic=66358.0

[prettybird]

Thanks for the response.

As I mentioned, I am in the US, and our office has a fair bit of interference in the 2.4GHz range. We're a large open-air office, and the 4 AP's are in each corner and situated about 100' from each other at the closest. Could such a (relatively) close proximity be an issue? There are quite a few freestanding structures throughout the office as well (some metal), so multiple locations have been necessary in the past for total coverage.

Are there any good reasons for why these things would simply power down like they've been doing? Even if it's an interference issue, why the reboots? It's beyond aggravating, and it leaves no hint as to why it happened. Just five minutes ago our "master" AP went down a few times in a row, nuking the connections of everyone in the office off and on for about 10 minutes.

1.17 was the latest firmware according to the official site, as far as I could tell. I tracked down the one you linked and will give it a shot.

[FurryNutz]

If you have a lot of congestion in your are due to other wifi neighbors, and leaving the DAPs on Auto Channel Scan will cause these disconnects as I presume they are encountering possible interference on channels being used and may not be able to find a clear channel to use.

I would have a look at the channel usage and see which channel is least used, 1, 6 or 11.
Also something that may help, if the DAPs have any power output levels, change this from HIgh to Medium or Low and test.

I put in a direct link to the FW at the Canada site.

[prettybird]

Thank you, I found the firmware earlier through Google -> Softpedia. The one on the Canada site seems identical.

Upgrades are done, so hopefully that helps things. No drops as of yet.

As for the reboots, I understand that interference can cause drops, but we're talking about hardware reboots here, as in the AP's are powering down entirely, ungracefully, and causing their clients to stall out and not reconnect elsewhere. It's a mess. I would think that it's faulty hardware, but it's happening to all four of our DAP-2695's at one point or another. What could explain that?

I'll try setting the channels manually. The "Channel Analyze" function is definitely helpful for this.

What about those other two issues? Non-functional NTP settings and super long AP handoffs? It's still weirding me out seeing 1970 as the system year, and walking from one room to another is still a roll of the dice even when the AP's are behaving.

[FurryNutz]

Not sure about the NTP and AP handoffs issues. I'll see if I can get some additional review on this.

NTP, is the feature not working in regards to UI operation or actual not getting NTP traffic from the WAN?

You might take one of the DAPs home with you if you can and test one out at home to see if NTP can be gained in a different environment.

What ISP Mfr and model modem do you have?
What Mfr and model main host router do you use that the DAPs are all connected to?

[prettybird]

The AP's can get to the internet without any trouble, and NTP services work just fine everywhere else on the network. The settings will save just fine, but no updates ever happen. Time remains at 1970 unless manually set, and of course this is wiped out when these things inevitably reboot.

Our ISP is Level3, and the network backbone is DAP-2695's --> Adtran NetVanta 1534 gig switch --> Adtran NetVanta 4430 router.

It seems that several reboots happened overnight, and again no logs are saved permanently so I have no idea what's going on or why. Do you have any insights into why these reboots are happening? I've dialed them down to 50% antenna output for now (kind of defeating their purpose) and am having to adjust the RSSI timeouts so we can actually maintain connections now. Even still, clients are being dropped and service remains unreliable.

I could take one home and try to get NTP working, but frankly that's the least of my concerns when the very basics aren't even working. I'm about to print some return labels and get rid of these things.

[FurryNutz]

Ok, the full on unit reboots I presume should not be happening. I would upgrade FW on 1 unit and test to see if the problem continues...

[drshock]

Couple of comments here from a fellow DAP-2695 A2 owner.

- On NTP this feature works fine for me, but you have to use the IP address rather than the name.   In my case I use a USA NIST server, 98.175.203.200 and set my USA timezone and enabled daylight savings time.    I've never had any problems with it since initial setup over a year ago.   
 
- I also use just one SSID across both radio bands.  In fact I share it across this model and the DAP-2553 B1 model in the same building.   But I do force under wireless settings them all to be WPA2 AES only on both radio bands.   I found once I did this I never had anymore connectivity issues.   Also forcing mixed 802.11ac + 802.11n on the 5Ghz, and 802.11n only on 2.4Ghz.  Supporting the obsolete 11a/b/g junk is just not worth the hassle.  Another important setting for Apple products is to force 20Mhz channel width on the 2.4Ghz band, none of their products support wide band channels https://support.apple.com/en-us/HT202068 .   This is not an issue on 5Ghz, go full auto wide there.  I had lots of Apple device issues with bouncing connections until I dropped wide band support on 2.4Ghz.
 
- I never found the AP Array feature to be stable, master and slave DAPs would reboot randomly.   D-Link support told me that it was not supported to mix model types in an AP Array, so my experience may differ from yours.  But once I disabled the array on all DAPs, the different model DAPs have worked well together for over a year.    I simply manually adjusted their placement and signal based on heat map app measurements on a MacBook doing walk thrus of the building floors.

- I found the aging out settings to negatively affect many client device types.  Constant connection drops and failure to smoothly transition between APs.  Some devices would just constantly bounce between connected and getting aged out.   So I ended up disabling this feature completely.  I did use the ACL RSSI feature and went with a 40% threshold with no problems.   
 
Enable the QoS feature under Traffic Control.   But be sure to enter the uplink/downlink settings based on your wired ethernet LAN on the screen before it, also under Traffic Control.

Leave the DAP to auto select channels on both radio bands, rather than manually setting them.

Also do not enable 802.11n preferred on 5Ghz or your 802.11ac clients will get steered to downgrade to 11n.   Definitely do enable this for 2.4Ghz band though.

Hope these tips help you get them going.   
   

[prettybird]

Drshock, thank you SO MUCH for taking the time to share these tips. I've changed up some of my settings accordingly, and it seems to be a big help. I'd already printed a return label and was going to ship these back, but will now see if they can actually remain stable for a day.

Thanks again!

[FurryNutz]

Glad you got them working better. Enjoy.