Prowess Data Unloader Complete Manual

PDU recovers MongoDB data. This documentation is organized by failure type and includes command and parameter references.

Scope

Source versions: 6.0.0 · 7.0.0 · 8.0.0

Linux

PDU for MongoDB 0.2.0

Offline file analysis

This documentation applies to Linux and source versions MongoDB 6.0.0, 7.0.0 and 8.0.0. Freeze source copies and launch PDU from a directory outside the sources. State and output use the launch directory without separate path settings. Confirm other versions and special layouts separately.

Common entry points

Recovery scenariosChoose a guide by instance, file, or accidental operation.Usage GuideReview configuration, execution order, and outputs.ReferenceFind commands, parameters, supported types, and frequently asked questions.

Recovery scenarios

Confirm the current database state, then select the matching procedure.

File and Directory Recovery

WiredTiger Files and Checkpoints

Read selected checkpoints from frozen sources, retaining catalog bindings and unresolved dependencies.

Recover from Corrupted InstanceRead the frozen WiredTiger data directory when the instance cannot start.MaterialdbPathcatalogOutputCheckpoint dataRecover from Corrupted Data FileProbe one WiredTiger file and export parseable records.MaterialWT fileversionOutputPhysical-version candidatesRecover Single DatabaseSelect one database for export from a complete WiredTiger directory.MaterialdbPathcatalogOutputCollection data and recordsRecover Metadata Relationships and Log PrefixesSalvage independent collection relationships when metadata is missing and build read-only recovery views from continuous WT logs.Materialcatalog / WT logscheckpoint LSNOutputIndependent relationships and read-only views

Historical Candidate Recovery

Collection History, HS and Pre-Images

Extract candidates from retained old values and source chains without automatically reconstructing business time.

Recover Deleted Records (Historical Candidates)Extract pre-delete data from retained historical versions and pre-images.Materialcollection historyHSOutputOld-version candidatesRecover Updated Records (Historical Candidates)Extract pre-update data from retained historical versions and pre-images.Materialcollection historyHSOutputOld-version candidatesRecover Old Versions from History StoreRead catalog-bound WiredTigerHS.wt with missing-page and modification-chain dependencies retained.MaterialWiredTigerHS.wtcatalogOutputOld values and missing-chain reasonsRecover Pre-Image Old DocumentsExtract retained old documents and filter operation ranges from oplog source chains.Materialpre-imageoplogOutputOld documents and attributionRecover Data at an Incident TimeSelect incident-time candidates using checkpoint, history-chain and original commit evidence.Materialcheckpoint / HScommit evidenceOutputTime-snapshot candidates

Mapped and Image Recovery

File Extents and Isolated Complete Blocks

Extract parseable content through explicit mappings or bounded scans, retaining unknown attribution.

Recover Data from Mapped FragmentsRead scattered pages, overflow and catalog dependencies through declared logical-file ranges.Materialimageextent mapOutputMapped records and dependenciesScan Isolated Complete Blocks in an ImageExtract complete WiredTiger blocks within explicit ranges, retaining unknown attribution and missing dependencies.Materialimagescan rangeOutputComplete-block candidates

Deep Fragments and Multiple Sources

Plaintext Bytes and Independent Source Cases

Retain complete-document candidates, field fragments and suspected structures separately, with additional sources and native resume.

Recover Plaintext Deep FragmentsExport complete BSON document candidates, independent field fragments and suspected structures separately.Materialfile / imagereadable bytesOutputDocuments/fields/structuresRecover Data from Multiple SourcesRegister directories, files and images independently, with added sources, failure isolation, pause and native reporting.Materialdirectories / files / imagesOutputIndependent results and summaries

Basic workflow

Preserve the original files, identify the problem, then carry out recovery on copies.

  1. 1

    Preserve recovery materials

    Copy dbPath, history store, pre-images and logs, and record the MongoDB version.

  2. 2

    Identify the failure type

    Separate instance damage, file damage, accidental DML, and dropped collections.

  3. 3

    Recover from a copy

    Choose the matching commands and record PDU's success and failure counts, original errors and checkpoints.

Reference

Find command syntax, parameter purpose, and version differences.

Command ReferenceSyntax and output descriptionsParameter ReferencePurpose and applicable conditionsSupported Data TypesSource formats and parsing conditionsFrequently Asked QuestionsAnswers to common recovery questions

Supported Data Types

NumericInt32 / Int64 / Double / Decimal128
Boolean and nullBoolean / Null / Undefined
Documents and identifiersString / Object / Array / Binary / ObjectId
Date and timeDate / Timestamp / Regular Expression
Advanced typesJavaScript / JavaScript with scope / MinKey / MaxKey
MongoDB 文档