ephemcache Helm values reference

Helm values reference table for the ephemcache application.

Key	Type	Default	Description
activeDeadlineSeconds	int	3600	Deadline for a single run, in seconds, so a wedged job cannot run forever
affinity	object	{}	Affinity rules for the pod
backoffLimit	int	0	Number of retries before the Job is considered failed. Zero, because the next scheduled invocation retries anyway.
command	string	"selftest"	Entrypoint subcommand. selftest checks the image and needs no database credential; run builds the cache for the current observing night.
database.passwordSecretKey	string	"usdf_mpc_postgres_password"	Key in this application’s Vault secret holding the database password, supplied to the pod as PGPASSWORD
database.url	string	""	SQLAlchemy DSN for the MPC replica, supplied as MPCDB. Empty uses the value baked into the image. Must not contain a username or password: it is passed to get-mpcorb.py as a command-line argument.
database.userSecretKey	string	"usdf_mpc_postgres_user"	Key in this application’s Vault secret holding the database username, supplied to the pod as PGUSER
failedJobsHistoryLimit	int	3	Number of failed Jobs to retain
global.host	string	Set by Argo CD	Host name for ingress
global.vaultSecretsPath	string	Set by Argo CD	Base path for Vault secrets
image.pullPolicy	string	"IfNotPresent"	Pull policy for the image. Use Always where the tag is mutable, such as a branch name.
image.repository	string	"ghcr.io/mjuric/lsst-gen-ephemcache"	Image to run, built by the lsst-gen-ephemcache repository
image.tag	string	"u-mjuric-ephemcache"	Tag of the image to run
logToFile	bool	false	Whether to also write each run’s log to a file under the output directory. Logs always go to stdout; this is a durable second copy that outlives the pod.
ncores	string	""	Value of NCORES, which bounds sorcha’s parallelism. Must match resources.limits.cpu: the scripts otherwise default it to nproc, which reports the node’s core count rather than the pod’s limit.
nodeSelector	object	{}	Node selector rules for the pod
outputDirMount.claimName	string	"sdf-group-rubin"	Name shared by the PVC, its storage class, and the pod volume
outputDirMount.enabled	bool	false	Whether to mount the shared filesystem that output is published to. When disabled, output goes to an emptyDir and is discarded with the pod.
outputDirMount.storageClassName	string	"sdf-group-rubin"	Storage class backing the claim
outputDirMount.subPath	string	"web_data/mpsky-data/test"	Path within that filesystem to expose. Only this subtree is visible to the container, so widening it is a deliberate change.
podSecurityContext	object	{}	Security context for the pod
resources	object	{"limits":{"cpu":"1","memory":"2Gi"},"requests":{"cpu":"1","memory":"2Gi"}}	Resource requests and limits. Sized for selftest; a real run needs far more, and must be raised together with ncores.
schedule	string	"0 */3 * * *"	Cron schedule for the cache build. Runs are idempotent and skip when the cache for the current observing night already exists, so this only has to be frequent enough to catch the night.
successfulJobsHistoryLimit	int	3	Number of completed Jobs to retain
suspend	bool	true	Whether to suspend the CronJob
tolerations	list	[]	Tolerations for the pod