Working with timestamps

Ideally, each CSC records the time at which the telemetry being reported was measured. The CSC timestamp field can be found by inspecting the topic schema from the SAL documentation or using the EFD client.

For example, for the MTMount.azimuth telemetry topic:

from lsst_efd_client import EfdClient
client = EfdClient("usdf_efd")

schema = await client.get_schema("lsst.sal.MTMount.azimuth")
schema

From which you see, among the other fields:

name        description                                               units      aunits   is_array

timestamp   Time at which the data was measured (TAI, unix seconds).  second     s        False

When the CSC timestamp is not present, you can use a SAL timestamp.

SAL timestamps are added automatically by SAL and have the private_ prefix. In particular, the private_sndStamp timestamp is the time at which the message was published by the CSC in TAI, which should be very close to the time at which the telemetry was measured. In fact, many telemetry topics rely on the private_sndStamp timestamp for that.

The private_sndStamp timestamp is also converted to UTC and recorded in the private_efdStamp field, which is used as the time index in the EFD.

Chronograf

Chronograf dashboards display time in UTC or in your local timezone. This can be changed by using the UTC/Local toggle in the Chronograf UI.

The function now() in InfluxQL returns time in UTC and can be used to query data relative to the current timestamp:

SELECT "actualPosition" FROM "efd"."autogen"."lsst.sal.MTMount.azimuth" WHERE time > now() - 1h

See Time Syntax for the time syntax when querying the EFD with InfluxQL.

EFD client

When querying the EFD with the EFD client, it is a good practice to convert all timestamps to UTC first.

Suppose you read the time of the start and end of an observation from the image header in TAI. Use the astropy.time package to convert time from TAI to UTC.

from astropy.time import Time

start_time = Time(header["DATE-BEG"], scale="tai").utc
end_time = Time(header["DATE-END"], scale="tai").utc

Then use the UTC timestamps in your EFD query:

actual_position = await client.select_time_series("lsst.sal.MTMount.azimuth",  ["actualPosition"], start_time, end_time)

See the EFD client documentation for other ways to query time ranges.