BUG - Spectral index maps not triggering

[pleasant-menlo]

Description of the issue

The ENA spectral index "spx" map products are intended to be produced in response to a new ENA intensity file being uploaded. However, they do not seem to be triggering currently.

Expected vs Actual behavior

Each spectral index job has a single ENA intensity map as input. When any new file with that descriptor is uploaded, we would like the spectral index job to run with the new file as input. Ideally this would happen quickly, just like ordinary daily files, but a delay of a few hours would be acceptable.

For example, see this dependency configuration:

sds-data-manager/sds_data_manager/orchestration/dependencies/imap_hi_dependencies.yaml

Lines 942 to 952 in b6ba928

	(l3, h45-spx-h-hf-sp-full-hae-4deg-6mo):
	partition: 6mo
	inputs:
	- source: hi
	data_type: l3
	descriptor: h45-ena-h-hf-sp-full-hae-4deg-6mo
	outputs:
	- source: hi
	data_type: l3
	descriptor: h45-spx-h-hf-sp-full-hae-4deg-6mo
	major_version: 1

A new file imap_hi_l3_h45-ena-h-hf-sp-full-hae-4deg-6mo_20260117_v001.0007.cdf was uploaded at 20260729 20:24:19, which should have triggered a Hi L3 h45-spx-h-hf-sp-full-hae-4deg-6mo job. No such job has been seen.

Product list

(hi, l3, h45-spx-h-hf-sp-anti-hae-4deg-1yr)
(hi, l3, h45-spx-h-hf-sp-anti-hae-6deg-1yr)
(hi, l3, h45-spx-h-hf-sp-full-hae-4deg-6mo)
(hi, l3, h45-spx-h-hf-sp-full-hae-6deg-6mo)
(hi, l3, h45-spx-h-hf-sp-ram-hae-4deg-1yr)
(hi, l3, h45-spx-h-hf-sp-ram-hae-6deg-1yr)
(hi, l3, h90-spx-h-hf-sp-anti-hae-4deg-1yr)
(hi, l3, h90-spx-h-hf-sp-anti-hae-6deg-1yr)
(hi, l3, h90-spx-h-hf-sp-full-hae-4deg-6mo)
(hi, l3, h90-spx-h-hf-sp-full-hae-6deg-6mo)
(hi, l3, h90-spx-h-hf-sp-ram-hae-4deg-1yr)
(hi, l3, h90-spx-h-hf-sp-ram-hae-6deg-1yr)
(hi, l3, hic-spx-h-hf-sp-full-hae-4deg-1yr)
(hi, l3, hic-spx-h-hf-sp-full-hae-6deg-1yr)
(lo, l3, l090-spx-h-hf-nsp-ram-hae-6deg-1yr)
(lo, l3, l090-spx-h-hf-sp-ram-hae-6deg-1yr)
(lo, l3, l090-spx-h-sf-nsp-ram-hae-6deg-1yr)
(lo, l3, l090-spx-h-sf-sp-ram-hae-6deg-1yr)
(lo, l3, l090-spxnbs-h-hk-nsp-ram-hae-6deg-1yr)
(lo, l3, l090-spxnbs-h-sf-nsp-ram-hae-6deg-1yr)
(ultra, l3, u45-spx-h-hf-sp-full-hae-2deg-1yr)
(ultra, l3, u45-spx-h-hf-sp-full-hae-2deg-3mo)
(ultra, l3, u45-spx-h-hf-sp-full-hae-2deg-6mo)
(ultra, l3, u45-spx-h-hf-sp-full-hae-4deg-1yr)
(ultra, l3, u45-spx-h-hf-sp-full-hae-4deg-3mo)
(ultra, l3, u45-spx-h-hf-sp-full-hae-4deg-6mo)
(ultra, l3, u45-spx-h-hf-sp-full-hae-6deg-1yr)
(ultra, l3, u45-spx-h-hf-sp-full-hae-6deg-3mo)
(ultra, l3, u45-spx-h-hf-sp-full-hae-6deg-6mo)
(ultra, l3, u90-spx-h-hf-sp-full-hae-2deg-1yr)
(ultra, l3, u90-spx-h-hf-sp-full-hae-2deg-3mo)
(ultra, l3, u90-spx-h-hf-sp-full-hae-2deg-6mo)
(ultra, l3, u90-spx-h-hf-sp-full-hae-4deg-1yr)
(ultra, l3, u90-spx-h-hf-sp-full-hae-4deg-3mo)
(ultra, l3, u90-spx-h-hf-sp-full-hae-4deg-6mo)
(ultra, l3, u90-spx-h-hf-sp-full-hae-6deg-1yr)
(ultra, l3, u90-spx-h-hf-sp-full-hae-6deg-3mo)
(ultra, l3, u90-spx-h-hf-sp-full-hae-6deg-6mo)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-2deg-1yr)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-2deg-3mo)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-2deg-6mo)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-4deg-1yr)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-4deg-3mo)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-4deg-6mo)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-6deg-1yr)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-6deg-3mo)
(ultra, l3, ulc-spx-h-hf-sp-full-hae-6deg-6mo)

[lacoak21]

I will look into this today.

[lacoak21]

Unfortunately I did not have time to look into this before I head off for vacation. I will try to pass off this ticket.

[pleasant-menlo]

@bryan-harter would we be able to view the Dagster logs for the Hi L3 h45-spx-h-hf-sp-full-hae-4deg-6mo product? We saw that there was an ENA map (imap_hi_l3_h45-ena-h-hf-sp-full-hae-4deg-6mo_20260117_v001.0007.cdf) uploaded last night that should have triggered that SPX job, but no job appears in the Processing Jobs table.

Do you have any ideas about why it did not trigger?

[bryan-harter]

I believe it might have to do with the map jobs taking longer than expected, Dagster timed out, though the file successfully completed. I'll make a fix today to make the time out longer