Logs showing 40 times per second errors connecting to ps.pndsn.com

1

Docker container sometimes crashes so i was trying to look into the HA logs (which were 1.8 gigs for 7 days) and found the following. Unsure how to track down what is trying to do this or why it’s trying 40 times per second 24/7. Google says it’s pubnub. Unsure what that is. Running on Synology Nas in docker if that matters. Any help would be greatly appreciated.

2025-11-14 15:42:26.564 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.566 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.567 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.569 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.570 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.573 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.574 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.575 WARNING (Thread-1 (_monitor)) [homeassistant.util.logging] Module pubnub is logging too frequently. 200 messages since last count
2025-11-14 15:42:26.577 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.577 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.579 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.580 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.582 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.583 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.585 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.585 ERROR (MainThread) [pubnub] Exception in subscribe loop: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]
2025-11-14 15:42:26.588 ERROR (MainThread) [pubnub] session.request exception: Cannot connect to host ps.pndsn.com:443 ssl:default [Could not contact DNS servers]

2
3

What hacs/add-ons and integrations are you using?
Anything chat related?

1000076626
10000766261224×1440 170 KB

4

Airthings
August
Bond
Denon
Ecobee
EspHome
Cast
Govee
Honeywell
Logitech Harmony
Philips Hue
Plex
Reolink
Smartthings
Tapo Cameras
TpLink switches

5

Maybe logitech.

6

A number of those vendors are powered by PubNub. The question is why can’t your docker container connect to DNS to resolve the ps.pndsn.com domain? I suspect the issue may be with the docker network config rather than the connection to pubnub?

7

Would the best solution be to disable each integration one by one and see ? Or is there a better way to improve the logging to actually say more information that the basic stuff it’s showing now ? Appreciate everyones help so far… !!

8

I concur with Todd, would be weird it just starts to happen all of a sudden. Check the network first.

And yes disable the add-ons one by one or the other way around: all at once and re-enable them one by one.

9

Faster to disable half. If it goes away it’s in that half if not it’s in the other. Turn them back on and pick half of the remaining ones…

Repeat cutting your set in half until you find it. Probably the second or third split.

10

Haha awesome is this a sorting algorithm you are using?

1 Like
11

It basically is - It’s how I used to Identify busted sections of an Active Directory Installation. You’re effectively cutting your choices in half zeroing in on the problem - mathematically unless you have a CRAPTON of things to sort you’ll rarely need more than 5-6 tries. The more things you have to check the more you eliminate in the early rounds.

12

Gonna ‘borrow’ this for future reference.

1 Like
13

I think it is like 35 years ago I first learnt that testing method.
Here it is called what could be translated to Successive Testing in English.

1 Like
14

Cutting things in half and then testing made this much faster. Thank you. So It looks like it’s Logitech Harmony and Honeywell. I simulated it by disabling the network and each of those would write those entries. I guess I’m still wondering why they would be trying so frequently if there was a network issue. Is that normal to have so many attempts, and why it would happen if there were no network issues as well…

1 Like
15

I believe the Logitech is using a very, very old PubNub SDK that retries a lot, vs. the newer PubNub SDKs that fallback more gracefully.

1 Like