Issues
- Physical backup with files that have reached remote storage limitsPBM-1490
- PiTR with physical backup doesn't recover dataPBM-1167Boris Ilijic
- PITR will stop saving the oplog if the process of copying the oplog from the backup failsPBM-1488
- Sharded cluster restoration failure in 2.8.0PBM-1493Boris Ilijic
- PBM backup fails with context deadline exceeded errorPBM-1401Jakub Vecera
- Add PBM support for Workload Identity AuthenticationPBM-1504
- Add pbm cancel-restore commandPBM-1330
- pbm log says "waiting for the oplog" when it's actually waiting for the backup to finishPBM-1321
- Provide documentation for executing physical backups using PBM/PSMDB container imagesPBM-1204radoslaw.szulgo
- Support timeseries in collection renamePBM-1424
- Error: get metadata: parse physical restore status: read ts from cluster.hb: strconv.ParseInt: parsing "": invalid syntaxPBM-1238Boris Ilijic
- Add proper error message when PBM agents aren't availablePBM-731Jakub Vecera
- Improve incremental backup nomination mechanismPBM-1451
- s3 credential leakPBM-960radoslaw.szulgo
- Resync does not always update bucket resulting in faulty list and backups to wrong bucketsPBM-1038
- Agents stop processing requests in case PBM runs with AzurePBM-1093
- pbm describe-backup doesn't report backup error when pbm status doesPBM-1116
- PBM backup upload to S3 fails when s3 bucket locked with modify or deletePBM-1125
- Fail of PITR slicer on one RS doesn't stop the whole PITR processPBM-1345Boris Ilijic
- Automatically raise the pbm user privileges during restorePBM-367
- Don't start backup/restore if agents are of different pbm versions.PBM-667
- restore with a different db namePBM-690
- Remote backup storages support Swift compatible storage
PBM-693 - Update status pbm-agent exitPBM-784
- PBM - Add support to BoostFSPBM-796radoslaw.szulgo
- Allow to exclude collection from PBM backupPBM-798
- Removing artifacts from storage if backup was unsuccessfulPBM-835
- pbm should print a friendly error when trying to execute physical backup in the inMemory enginePBM-843
- Add status report mechanism for all commandsPBM-950
- Physical PITR (log-based incremental physical backups)PBM-973radoslaw.szulgo
- Allow the pbm logs command to retrieve complete log from pbm-agents during physical restorePBM-1092radoslaw.szulgo
- Custom S3 TLS Certificate support in PBMPBM-1525radoslaw.szulgo
- PBM mongo backup use too much memoryPBM-1126
- PBM pre-check leads to a wrong understanding.PBM-1206Jakub Vecera
- pbm delete-backup does not wait for backup to delete, does not queue up additional requestsPBM-1227radoslaw.szulgo
- PiTR fails after physical restorePBM-1265Resolved issue: PBM-1265
- Add support for logical restore of sharded timeseries collectionsPBM-1295radoslaw.szulgo
- Provide post-backup validation functionalityPBM-1322
- SPIKE - investigate the possible solution for s3 credential leakPBM-1392
- Subcommand to verify backup filesPBM-1434radoslaw.szulgo
- Selective point-in-time restore for sharded collectionsPBM-1437
- Data isn't removed from storage after cancelling logical backup in sharded clusterPBM-1446
- Physical restore fails with arbitersPBM-1449Boris Ilijic
- pbm backup names - custom name / without special charsPBM-1462
- Fix backup type in "another operation in progress" error messagePBM-1464
- Automate using a physical backups to seed a lost node.PBM-1172radoslaw.szulgo
- Extra slash symbol in bucket/prefix options breaks PBM ability to save or discover backups on storagePBM-1472
- Physical restore assumes a functioning clusterPBM-1335Boris Ilijic
- PBM cannot restore/resync from a read-only datastore, improve to allow this to work even if some restrictionsPBM-1480
- PBM SDK: list oplog rangesPBM-1225Jakub Vecera
50 of 55
Physical backup with files that have reached remote storage limits
General
Escalation
General
Escalation
Description
Environment
None
Created February 11, 2025 at 6:57 AM
Updated April 1, 2025 at 11:59 AM
Activity
radoslaw.szulgoMarch 27, 2025 at 3:00 PM
Feedback from the customer to increase the prior: our primary cluster backing all tooling that supports all internal analysts, researchers, support engineers and executive staff won’t be regularly backed up to a location supported by our operations team. Getting a regular incremental backup set up with PBM is critical to us.
radoslaw.szulgoFebruary 11, 2025 at 9:40 AM
Shouldn’t the selective backup be used to split the size of the backup into separate buckets?
The problem
Maximum size of AWS S3 object is 5TB so if trying uploading files larger than 5TB PBM fails with the EntityTooLarge error
Proposed solution
PBM should split the file on EntityTooLarge error, store this information in the metadata, and merge everything back during restore
Acceptance criteria
applies only on physical backups
need to have configuration parameter of the file size
QA and docs
need to have configuration parameter of the file size