We are running a single machine GeoEvent 10.6.1 (10.6.1.9310, 10.6.1.9310 - 10.6.1 Patch 1, 10.6.1.9301 - 10.6.1 Hotfix 1) and have a situation where the data is old by the time it is being written to the db. We set the logs to debug and noticed that the timestamp in the outgoing message was at times an hour old against the time noted in the logs, shown below.
2019-10-08T15:35:17,660 | DEBUG | HttpRequest Worker Thread: https://xxx/gissite/rest/services/xxx/FeatureServer/0/updateFeatures | wire | 53 - com.esri.ges.framework.httpclient - 10.6.1 | http-outgoing-5061669 >> "f=json&token=xxx&features=%5B%7B%22geometry%22%3A%7B%22x%22%3A145.28335%2C%22y%22%3A-38.11364%2C%22spatialReference%22%3A%7B%22wkid%22%3A4326%7D%7D%2C%22attributes%22%3A%7B%22radioid%22%3Axxx%2C%22timestamp%22%3A1570505632872%2C%22infotype
log time 2019-10-08T15:35:17,660 vs timestamp of 1570505632872 (2019-10-08T14:33:53)
I don't think there is a delay between the input and output connector and would like to check that there is no backlog prior to the input connector. I have checked the :6443/arcgis/admin/system/platformservices status
{"configuredState": "STARTED",
"details": [{
"machine": "xxx",
"realtimeState": "STARTED"}]}
and health - the result of which I am assuming is ok
{"xxx": {}}
The GeoEvent Monitor shows the same rate and "time since last" for the input and output.
The machine CPU and memory is low at around 20% and 38% respectively.
Does anyone have any tips on where else I could look for clues as to the internal performance of this GeoEvent service please? TIA.