On 4/10/25 14:21, Craig Leres via OvmsDev wrote:
I'm now testing 3.3.004-313-g78a49a59 (which is older than 3.3.004-323- gbc1f75c9, yet another git annoyance for me that the hashes are not chronological...) But I think it has one change that 3.3.004-313- g78a49a59 does not.
It looks like 78a49a59 is the source of my hologram usage increase. I see these: + m_updatetime_keepalive = 29*60; + m_legacy_event_topic = true; But m_legacy_event_topic doesn't look like it has an impact (mg_mqtt_publish() still gets called). However, this worries me: + [server.v3] updatetime.keepalive -- Max number of seconds to allow the MQTT connection to be idle before + sending PINGREQ. Should be set slightly shorter than the network's + NAT timeout and the timeout of your MQTT server. If these are unknown + you can use trial and error. Symptoms of keepalive being too high are + a lack of metric updates after a certain point, or "Disconnected from + OVMS Server V3" appearing in the log. In previous releases this used + the Mongoose default of 60. New default value 1740 aligns with + observed Hologram timeout of 1800. but, again, not clear how it can drive up cellular usage when I'm on wifi. (Unless there is some kind of wifi-sleep mode happening that causes cellular to be used?) So maybe this is just triggering a previously latent bug? Anyway, I don't see how to disable this short of commenting code out. Craig OVMS# config list server.v3 server.v3 (readable writeable) port: server: example.net tls: yes topic.prefix: updatetime.connected: updatetime.idle: user: fred