Skip to main content

CRD reference

Gryvia defines 50 custom resources in the gryvia.io API group (version v1alpha1, alpha: fields may change between releases). The last column names the operator that registers a controller for the kind; none means no controller in operators/*/main.go reconciles the kind: the CRD, gateway routes or dashboard pages may exist, and other components may read it (for example GryviaGpuSku), but nothing acts on its spec. Full schemas are in crds/; kubectl explain <kind>.spec works on any cluster with Gryvia installed.

KindResourceScopeControllerRequired spec fieldsOther spec fields
GryviaAIJobgryviaaijobsNamespacedai-operatorgpus, image, typeaffinity, args, command, distributed, env, gpuType, imagePullPolicy, model, …
GryviaAuditgryviaauditsClusternonenoneanomalyDetection, compliance, events, reporting, retention, scope
GryviaAutoPolicygryviaautopoliciesNamespacednetwork-intelligencenoneapprovalRequired, excludeServices, learningWindow, mode, targetNamespaces
GryviaAutoScalergryviaautoscalersClusternonegpuType, maxNodes, queueRefadvanced, costControls, minNodes, nodeProvider, scaleDownPolicy, scaleUpPolicy
GryviaAutoTunergryviaautotunersNamespacednonejobTemplate, maxTrials, objective, parameterSpace, searchAlgorithmashaConfig, earlyStoppingRounds, parallelism
GryviaBenchmarkgryviabenchmarksNamespacednonetarget, typebaseline, custom, gpuMemory, ioThroughput, mlperf, nccl, schedule
GryviaBudgetgryviabudgetsClusternoneperiod, scopealerts, enforcement, limits, priority, rollover
GryviaChargebackgryviachargebacksClusternonenoneallocationModel, costCenters, mode, period, pricing, reports
GryviaCheckpointGuardgryviacheckpointguardsNamespacedai-operatorcheckpointPolicy, jobSelectormonitoring, restore, validation
GryviaCostPredictorgryviacostpredictorsClusterquota-operatornonealternatives, historicalData, integration, models, pricing
GryviaDRTestgryviadrtestsClusternonetypeapprovalRequired, backupRestore, chaosEngineering, dataIntegrity, failover, fullDrill, notifications, rpoRto, …
GryviaDatasetgryviadatasetsClusternonesourceaccess, cache, description, license, statistics, tags, type, version, …
GryviaFabricSignalgryviafabricsignalsNamespacednonenonejobRef, observeOnly
GryviaFederationgryviafederationsClusternonenoneclusters, costManagement, distribution, failover, loadBalancing, resourceSharing
GryviaFlowPolicygryviaflowpoliciesNamespacednetwork-intelligencenoneaction, destination, intent, priority, protocol, source
GryviaGPUSharingPolicygryviagpusharingpoliciesClusternonestrategyfractionalGPU, mig, nodeSelector, priority, qos, tenantQuotas, timeSlicing
GryviaGpuMemoryOptimizergryviagpumemoryoptimizersClustergpu-operatorscopeinferencePacking, oomPrevention, rightSizing
GryviaGpuNodegryviagpunodesClustergpu-operatorgpuCount, gpuType, nodeNamebandwidth, computeCapability, drivers, healthCheck, interconnect, labels, memoryGB, rdma, …
GryviaGpuSkugryviagpuskusClusternonegpuType, hourlyRatecurrency, description, enabled, gpusPerUnit, spotDiscount
GryviaHealthCheckgryviahealthchecksClusternonetargetchecks, onFailure, remediation, schedule
GryviaInferenceInsightgryviainferenceinsightsNamespacednetwork-intelligencenoneanalysisWindow, targetService
GryviaInferenceServicegryviainferenceservicesNamespacednonebackend, modelRefargs, autoscaling, canary, gpuCount, gpuType, healthCheck, image, replicas, …
GryviaJobHookgryviajobhooksClusternoneaction, triggercondition, failurePolicy, retry, selector
GryviaLiveExperimentgryvialiveexperimentsNamespacedai-operatorcomparison, jobsdescription, notifications, strategy
GryviaMetricgryviametricsClusternonename, sourcedescription, labels, retention, thresholds, type, unit, visualization
GryviaModelLineagegryviamodellineagesNamespacedai-operatormodelcompliance, provenance
GryviaModelRegistrygryviamodelregistriesNamespacednoneartifacts, modelName, versionautoServe, description, metadata, servingConfig, source, stage
GryviaNetworkgryvianetworksClusternetwork-operatornetworkTypemtu, nodeSelector, rdma, sriov, targetNamespace
GryviaNetworkAnomalygryvianetworkanomaliesNamespacednetwork-intelligencenonealertWebhook, autoMitigate, detectionRules, targetService
GryviaNetworkCostgryvianetworkcostsNamespacednetwork-intelligencenonecostCenters, costPerGB, reportingInterval, targetNamespaces
GryviaNodeFabricgryvianodefabricsClusternonemeasuredAt, nodeName, scoreDeltaexpiresAt, reasons, ttlSeconds
GryviaPrioritygryviaprioritiesClusternonevaluedescription, preemptionPolicy, quotaOverride, sla
GryviaQuotagryviaquotasClusterquota-operatorgpuQuota, namespaces, teambudget, network, priority
GryviaQuotaPolicygryviaquotapoliciesClusternonenonealerts, allocation, enforcement, hierarchy, limits, scope, timeBased
GryviaReservationgryviareservationsClusternoneowner, resources, schedulebilling, guarantees, notifications
GryviaRetryPolicygryviaretrypoliciesClusternonenonebackoff, budget, circuitBreaker, maxRetries, noRetryOn, resourceAdjustment, retryOn
GryviaSLAgryviaslasClusternonescope, tieravailability, failureHandling, monitoring, performance, resources, support
GryviaSecurityPolicygryviasecuritypoliciesNamespacednetwork-intelligencenonealertWebhook, autoBlock, detectionRules, targetNamespaces
GryviaServiceGraphgryviaservicegraphsNamespacednetwork-intelligencenonedepth, includeExternal, namespaces, refreshInterval
GryviaStoragegryviastoragesClusterstorage-operatorbackend, capacity, endpointcredentials, iops, mountOptions, performance, protocol, quotas, rdma, storageClass, …
GryviaTemplategryviatemplatesClusternonecategory, defaultsdescription, parameters, tags
GryviaTenantgryviatenantsClusterquota-operatornoneallowedSkus, billing, description, displayName, governance, jobDefaults, members, networkPolicy, …
GryviaTraceSessiongryviatracesessionsNamespacednetwork-intelligencenonecaptureHeaders, duration, filters, level, namespace, service
GryviaTrafficInsightgryviatrafficinsightsNamespacednetwork-intelligencenonemetrics, namespace, service, window
GryviaTrainingInsightgryviatraininginsightsNamespacednetwork-intelligencenoneanalysisWindow, metrics, targetJob
GryviaTrainingProfilergryviatrainingprofilersNamespacedai-operatortargetanalysis, metrics, output
GryviaTrainingTimeMachinegryviatrainingtimemachinesNamespacedai-operatorsourceJobforks, retention, timeline
GryviaUsageRecordgryviausagerecordsNamespacedquota-operatorcost, final, gpuHours, gpus, job, jobUID, rate, start, tenantcurrency, end, gpuType, sku
GryviaWorkflowgryviaworkflowsNamespacednonestepsparameters
GryviaWorkspacegryviaworkspacesNamespacednonetypecpuLimit, cpuRequest, env, gpuCount, gpuType, idleTimeoutMinutes, image, maxLifetimeHours, …