Example:
Thread 1 writes:
time=2023-01-01T00:00:00Z
a=1 (tag)
b=1 (tag)
c=1 (field)
d=1 (field)
Thread 2 writes:
time=2023-01-01T00:00:00Z
a=1 (tag)
b=1 (tag)
e=2 (tag)
f=2 (field)
So the time and the tags are the same but the fields differ.
Expected result
If I read the record, I would expect this:
time=2023-01-01T00:00:00Z
a=1 (tag)
b=1 (tag)
c=1 (field)
d=1 (field)
e=2 (tag)
f=2 (field)
Is this assumption correct or will this end up in an error?
Thank you very much
A series is defined by the unique combination of measurements, tag key value pairs, and field keys.
You can learn more here
so if you wrote the following line protocol which corresponds to what you have
measurement, a=1,b=1,e=2 f=2 time1
measurement, a=1,b=1,e=2, c=1,d=1,f=2 time1
You’d get the following series:
measurement, a=1,b=1,e=2 f=2 time1
measurement, a=1,b=1,e=2, c=1 time1
measurement, a=1,b=1,e=2, d=1 time1
measurement, a=1,b=1,e=2, f=2 time1
But the first and the last one are the same so you’d be overwriting the first point (even if your f field value was different).
So you’d end up with a total of 3 series.
To be precise if you query with Flux,
(I’m assuming you’re using Flux and OSS),
You’d get 3 tables returned for each series which would each graph one line.
does that help?
This might also be useful
Please let me know if my version assumption was wrong
Thank you, @Anaisdg. I’m not using flux.
My use case is to write to the exact same measurement at exact the same time but with different fields. I updated the example:
Thread 1 writes:
time=2023-01-01T00:00:00Z
a=1 (tag)
b=1 (tag)
c=1 (field)
d=1 (field)
Thread 2 writes:
time=2023-01-01T00:00:00Z
a=1 (tag)
b=1 (tag)
e=2 (field)
f=2 (field)
What happens if two threads in the InfluxDB instance write to the storage at exactly the same time? Would this lead to an error or is InfluxDB capable of resolving this?
I would expect one measurement with the fields c,d,e,f set, as the timestamp and all tags are equal.
Thank you.
