#influxdb

that is for selecting an exact data point iirc

what i want is all records where sequence_number > y

but you know that sequence number is only used to differentiate amongst points with the exact same timestamp, right?

no i didnt

fiorix: so typically all your timestamps will have sequence number 0, unless you have multiple datapoints on the same timestamp

i didn't know that... so perhaps my sequence numbers are all zeroes because i have like 1 data point every other second

although that's not what the front-end shows.. those data points do have distinct sequence_number

how do you populate the db?

who sets the sequence numbers?

i don't.. but they do show up in queries

all i set in the db are the 2 columns i need.. but when i query i get 4 columns, as expected.. time, sequence_number, x, y

what is wrong with this query? delete from engines.tyland.system where time = time1407129391352

what is wrong with this query? delete from engines.tyland.system where time = 1407129391352

the second one

i get this error Delete queries can't have where clause that don't reference time

don't you need a s or ms or something at the end

why would I need that?

crosbymichael: you can only do < and > comparisons, not exact match

pauldix: why?

i need to delete a bad data point

Any recommendations on protobuf_timeout and protobuf_heartbeat in EC2? Currently using default of 200ms and I'm getting heartbeat timeouts for at least one node out of three every minute or so. All three are in same Region, but split between two AZs

crosbymichael: just what we have implemented at the moment

you could do where time > 1407129391351s and time < 1407129391353s

so what do you suggest I do to delete an exact data point?

assuming you don't have other data points in that series in the same time range