Unexpected high ingress

Hi everybody,
We are new to influxDB and are using influxDB as an alternative for Azure-TSI.
Our first test however show an unexpected high ingress count in the usage page of the influxDB portal. The ingress can be as high as 600 MB/hour while the network traffic measured with Wireshark towards the influxDB cluster is 72MB/hour. We are puzzeled by this huge difference.
Because this unexpected high amouth of ingress traffic will result in unexpected high usage costs, help or tips are much appriciated!

Regards,
JohanK
The Netherlands

@johankleuskens The ingress usage statistic measures uncompressed data that gets written to the database (as noted here). I’m guessing that your data is compressed when you send it over the network with your write request. This would explain the difference.

@scott We are using the InfluxDB 3 C# .NET Client library and looking at its source code, it looks like gzip is on by default. Indeed, this could explain the difference.

I will get back on this asap.

I should mention that gzip’ing data is a recommended best practice (which is why the client enables by default). With the rate limits in place on InfluxDB Cloud Serverless, this can be tricky knowing how much data is actually getting written to the database until after it’s written unless you’re somehow tracking the size of uncompressed data client-side.

After disabling compression, we concluded the amount of ingress trafic roughly matches the amount of traffic measured by Wireshark. Thanks to @scott for helping us with this issue.