uniFi-WiFI-AP: discussions

Jp. I am out of town. Don't know when I can check it.

Are these regular ibeacons or bleconnect...?


Sent from my iPhone using Tapatalk

No worries. When you have a chance.

This is not pibeacon related, but with the UniFi.WiFi-AP plugin. Again, up -> down, work immediately and down -> up the same. The problem is that down - > expired almost never happens. The device gets stuck in the down status.

JP

Ok sorry was on my phone and with Jetlag
Could you post the device def. = what are the parameters.


Sent from my iPhone using Tapatalk

Here you go.

Ok I believe if it is a fast down device it does not go to expire. One of the things left to do


Question is should it use the regular expire time(90 sec) or the fast time(10sec) to do from down to expired.
Any suggestion?

Sent from my iPhone using Tapatalk

If you have fastime setup, it should use that timeout (10sec) otherwise, it should use the default (90sec)

JP

Karl,

Just in case it helps for when you have time, I just noticed that the same device in fingscan also never goes from down to expired either. This is the fingscan device utility log:

id strftime('%Y-%m-%d-%H:%M:%S',ts,'localtime') status
---------- -------------------------------------------- ----------
7806 2017-02-18-08:52:19 up
7807 2017-02-18-08:53:04 down
7808 2017-02-18-08:53:15
7809 2017-02-18-08:53:20 up
7810 2017-02-18-08:53:40 down
7811 2017-02-18-08:54:11
7812 2017-02-18-08:58:32
7813 2017-02-18-08:58:32
7814 2017-02-18-09:02:51 up
7815 2017-02-18-09:16:21 down

Maybe not related but interesting that the same device in both plugins (fingscan and UniFi-WiFi) behaves the same. Something specific to the device? Is an iPhone 7.

JP

I have added the code for fastDown devices for UniFi. Just want to try it out remotely before I post.



Sent from my iPhone using Tapatalk

https://www.dropbox.com/s/tzqfm3bbysado6k/uniFiAP-v-7-2-32.zip?dl=1 v 7-3-32

roquej:

if you could be so kind and test the attached unifi version. It should go to expired for fastDown devices... BUT it is not really tested, the change was simple though. If you ok it I will post it in the download section.


As for your fingscan comment, I will need to test this. The code in fingscan and unifi is completely different.. Fingscan has a kind of "fastDown" if you use the ASUS router mode. But otherwise no.
And looking at the SQL dump it should switch to expired. Will look at that when I am home.


Karl

Karl,

The Fast Down seems to be working. Thanks!

JP

Karl,

For when you get back, the fast down is working, but is flaky. Devices sometime don't even go down/expired, even when they are not in the network. Example:

id strftime('%Y-%m-%d-%H:%M:%S',ts,'localtime') status
---------- -------------------------------------------- ----------
25597 2017-02-23-13:14:54
25598 2017-02-23-13:14:54
25599 2017-02-23-13:15:26 down
25600 2017-02-23-13:15:28 up
25601 2017-02-23-13:15:29
25602 2017-02-23-13:15:30
25603 2017-02-23-13:15:55
25604 2017-02-23-13:16:13
25605 2017-02-23-13:16:15
25606 2017-02-23-13:23:22

JP

FastDown only reacts to a active message from the ap. If the device simply disappears without message it does not react. You should use optimized instead. It also has an expiration timer.


Sent from my iPhone using Tapatalk

Karl,

Very flaky stuff going on. In either "optimize" or "wait", device goes up to expired in very unpredictable ways. Even with wifi on, it just bounces constantly. It also goes expired when wifi on and stays there until I reset the wifi.

JP

I noticed this hitting my Event log about 1 every minute. Two days ago I added Unifi 24 Port Switch and USG. Not really sure when this started but had to be in the past few days.

Apr 21, 2017, 3:36:21 PM
SQL Logger Error exception trying to insert row into table device_history_515195515 ("UniFi-28:10:7b:29:34:a6")
SQL Logger Error database disk image is malformed
SQL Logger Error INSERT INTO device_history_515195515 ("ghz", "mac", "sw_ip_port", "agedhcp", "ageswitch", "associatedwifi", "connectwifi", "created", "groupmember", "idletimewifi", "inlistdhcp", "inlistswitch", "inlistwifi", "ipnumber", "laststatuschange", "lastup", "name", "noisewifi", "powermgmtwifi", "rx_tx_ratewifi", "signalwifi", "statewifi", "status", "updownsetting", "uptimedhcp", "uptimeswitch", "uptimewifi", "vendor") VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s);
SQL Logger Error exception trying to insert row into table device_history_515195515 ("UniFi-28:10:7b:29:34:a6")
SQL Logger Error database disk image is malformed
SQL Logger Error INSERT INTO device_history_515195515 ("ghz", "mac", "sw_ip_port", "agedhcp", "ageswitch", "associatedwifi", "connectwifi", "created", "groupmember", "idletimewifi", "inlistdhcp", "inlistswitch", "inlistwifi", "ipnumber", "laststatuschange", "lastup", "name", "noisewifi", "powermgmtwifi", "rx_tx_ratewifi", "signalwifi", "statewifi", "status", "updownsetting", "uptimedhcp", "uptimeswitch", "uptimewifi", "vendor") VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s);

this looks like your SQL database is corrupt.

when you restart indigo does it give any indication that SQL logger has problems?

could you try indigo utilities and run some report on any device you like (print sql report at the bottom) it might show you that it has problems.

you should get similar error messages not just from unifi tables but from other devices that change and cause an entry into the SQL database.

you might need to reset your database or use the backup from 2 days ago. .. time machine might not create a usable backup of your sqlite db. (use indigo-utilities to do a copy of the db .. then time machine will be able to make a backup copy of that copy... time machine takes too long do make a copy, while indigo still write to the DB)

Karl