Since about a week, my hass.io on a RPi3 has been almost completely unresponsive on my two mobile devices, while it is still as snappy as ever on my PCs. I have two different mobile devices logged into HA, a 7" Lenovo Android tablet and my phone running Ubuntu Touch. They are both using wifi, but I have also tried with mobile data on my cell phone. My tablet is terribly unresponsive, but seems to update clock etc. with maybe half hour intervals.
The PCs, which work just fine with HA, are both running Ubuntu 16.04 and Firefox. One with wired network and the other one on wifi. I also tried Chrome on one of them, and it also worked fine.
What could be behind this very odd behavior on my mobile devices? I haven’t found anything peculiar in the log file except quite a few of these: 2019-06-05 17:14:02 ERROR (MainThread) [homeassistant.components.websocket_api.http.connection.1398754960] Client exceeded max pending messages [2]: 512
Mobile is unresponsive after restarting the tablet/phone? Are other apps slow? My Fire HD8 tablet chugs on HA after leaving it running for more than a day. I run a slew of custom Lovelace cards on my instance and I see Google Chrome taking up gigabytes of memory until I close Home Assistant.
You can try disabling certain custom Lovelace cards that might be CPU intensive. I know my cards tend to refresh more than they should…
I upgrade to 0.94 today and got a notification about custom-useful-markdown-card being deprecated. I switched to the new markdown-mod instead. After that, my mobile devices are responsive again.
I spoke too soon. It seems my mobile units, especially the android tablet is having trouble updating. My database quickly builds to over 1 GB and one error I am getting frequently is this:
Error in database connectivity: (sqlite3.OperationalError) no such table: events
[SQL: INSERT INTO events (event_type, event_data, origin, time_fired, created, context_id, context_user_id) VALUES (?, ?, ?, ?, ?, ?, ?)]
[parameters: ('state_changed', '{"entity_id": "sensor.hank_hkzw_so05_smart_plug_power", "old_state": {"entity_id": "sensor.hank_hkzw_so05_smart_plug_power", "state": "2.26", "attrib ... (742 characters truncated) ... :00", "last_updated": "2019-06-16T15:06:41.319058+00:00", "context": {"id": "a80117b6dcf14598b8a0a60bf74335ad", "parent_id": null, "user_id": null}}}', 'LOCAL', '2019-06-16 15:06:41.319167', '2019-06-16 18:24:49.527837', 'a80117b6dcf14598b8a0a60bf74335ad', None)]
(Background on this error at: http://sqlalche.me/e/e3q8). (retrying in 3 seconds)
I will try and exclude all sensors from the recorder now, to see if anything improves.