Chart values - ABBYY Documentation

Every value of the fineparser-enterprise chart, with its default.

images

Value	Default	Description
`images.pullSecrets`	`[]`	Image pull secrets.
`images.<component>.image`	The AWS Marketplace repository	Image repository for `orchestrator`, `worker`, and `ingest`.
`images.<component>.tag`	The chart’s version	Image tag.
`images.<component>.pullPolicy`	`IfNotPresent`	Pull policy. Must be `IfNotPresent` for the worker when `worker.prepull.enabled` is `true`.

serviceAccount

Value	Default	Description
`serviceAccount.create`	`true`	Create the service account the orchestrator and workers run under.
`serviceAccount.name`	`""`	Its name. Empty uses `<fullname>-orchestrator`. When `create` is `false`, set it to the existing account.
`serviceAccount.annotations`	`{}`	Annotations for a chart-created account.

awsMarketplace

Value	Default	Description
`awsMarketplace.region`	`""`	The region workers register in. Set only if `AWS_REGION` is not injected.

scaling

See Scaling.

Value	Default	Description
`scaling.minWorkers`	`0`	Workers kept running when the queue is empty.
`scaling.maxWorkers`	`20`	The most workers the pool runs.
`scaling.minimumThroughputSeconds`	`120`	How soon queued work should get capacity, in seconds.
`scaling.defaultUnitSeconds`	`30`	Starting estimate of the time one unit takes, in seconds.
`scaling.tickMs`	`15000`	How often the pool size is checked, in milliseconds.
`scaling.scaleDownCooldownMs`	`300000`	How long the queue must call for fewer workers before the pool shrinks, in milliseconds.

orchestrator

Value	Default	Description
`orchestrator.workLeaseSeconds`	`300`	How long a worker has to finish a unit before it is given to another worker.
`orchestrator.workMaxAttempts`	`3`	Attempts per unit before its job fails.
`orchestrator.workReclaimInterval`	`60`	How often expired units are returned to the queue. The default checks every 6 seconds.
`orchestrator.jobStartTimeoutSeconds`	`3600`	How long an accepted job may wait to start before it fails with `never-started`.
`orchestrator.jobRetentionSeconds`	`86400`	How long a finished job’s status is kept. `"infinity"` keeps it forever.
`orchestrator.terminationGracePeriodSeconds`	`10`	Pod grace period.
`orchestrator.storage.size`	`2Gi`	Size of the state volume.
`orchestrator.resources.requests`	`cpu: 1`, `memory: 1G`	Resource requests. No limits are set.

worker

Value	Default	Description
`worker.predefinedProfile`	`DocumentConversion_Normal`	Default predefined profile. See Recognition settings.
`worker.userProfile`	`""`	Default user profile, as a multi-line string.
`worker.terminationGracePeriodSeconds`	`600`	Time a stopping worker has to finish its unit.
`worker.resources.requests`	`cpu: 1`, `memory: 1G`	Resource requests per worker. No limits are set.
`worker.prepull.enabled`	`false`	Pre-pull the worker image on new worker nodes. See Scheduling.
`worker.prepull.taintKey`	`startup-taint.cluster-autoscaler.kubernetes.io/fineparser-image`	The startup taint worker nodes register with.

ingest

Value	Default	Description
`ingest.replicas`	`2`	Number of ingest pods.
`ingest.pageSplitThreshold`	`25`	PDFs with more pages are split into parts of this many pages.
`ingest.maxUploadSize`	`512Mi`	Largest accepted request.
`ingest.scratchSize`	`2Gi`	Scratch space per ingest pod. About twice `maxUploadSize`.
`ingest.cors.allowedOrigins`	`""`	Comma-separated origins allowed to call the API from a browser.
`ingest.service.type`	`ClusterIP`	Service type.
`ingest.service.port`	`8080`	Service port.
`ingest.resources.requests`	`cpu: 500m`, `memory: 500M`	Resource requests per pod. No limits are set.

storage

See Storage, EFS file system, and S3 output bucket.

Value	Default	Description
`storage.inProcess.efs.fileSystemId`	Placeholder	The EFS file system. Must be set unless `existingClaim` is.
`storage.inProcess.size`	`10Gi`	Requested size of the in-process claim.
`storage.inProcess.storageClass.name`	`""`	Name of the chart-created EFS StorageClass. Empty uses `<fullname>-in-process`.
`storage.inProcess.storageClass.reclaimPolicy`	`Delete`	Its reclaim policy.
`storage.inProcess.storageClass.parameters`	`uid`, `gid` `65534`, `directoryPerms` `755`	Extra EFS CSI driver parameters.
`storage.inProcess.existingClaim`	`""`	Use this claim instead of creating one.
`storage.output.s3.bucket`	Placeholder	The S3 bucket. Must be set unless `existingClaim` is.
`storage.output.existingClaim`	`""`	Use this claim instead of creating one.
`storage.s3.region`	`""`	The bucket’s region.
`storage.s3.mountOptions`	`[allow-delete, allow-other, uid=65534, gid=65534]`	Mountpoint mount options. Setting the list replaces the default.

storageClass

Value	Default	Description
`storageClass.create`	`false`	Create an encrypted `gp3` StorageClass for the orchestrator’s state.
`storageClass.name`	`""`	StorageClass for the orchestrator’s state. Empty uses the default class. Required when `create` is `true`.
`storageClass.provisioner`	`ebs.csi.aws.com`	Provisioner of the created class. `ebs.csi.eks.amazonaws.com` on EKS Auto Mode.

Scheduling

orchestrator, ingest, and worker each take nodeSelector, tolerations, affinity, and podAnnotations, all empty by default. See Node groups and scheduling.