Release preparation v1.153

Product Discussions STORJLINGS
1

New release v1.153 is already deployed on QA Satellite
Here is a list of changes:

Changelog

General

  • a2c647b cmd/tools/om-license: new tool to grant OM licenses to active users
  • 5528a6c all: use atomic types where possible
  • d6b15df all: use reflect.TypeFor instead of reflect.TypeOf
  • 22ce77e all: use strings.CutPrefix/Suffix

Satellite

  • d691d0f satellite/admin: add role-based access control for OIDC authentication
  • 96de797 satellite/satellitedb: add partner column to registration_tokens table
  • 63ca6d4 satellite/metabase: remove SkipPendingObject flag
  • 43674ed satellite/console: allow project invitation on closed registration satellite
  • 9dedeca satellite/metabase: add checksum support to UpdateObjectLastCommittedMetadata
  • ed6189c satellite/metabase: update ListObjects to return checksum info
  • acb7743 satellite/metabase: update object iterators to return checksum info
  • 48a8bf8 satellite/metabase: update FinishCopyObject to set checksum info
  • db0eb29 web/satellite: fix encryption type shown as Self-Managed after project creation
  • 9509b1c satellite/{console,web}: updates to support free trial logic in the white labeled UI
  • e34adc5 satellite/{console,web}: enable billing features for white labeled console
  • 3f2317f web/satellite: remove minimum charge banner
  • 6eed8e5 satellite/console: send user created notifications
  • 1530a02 web/satellite: resolve ‘circular chunk’ build issue
  • 11f1573 satellite/metabase: add checksum support to FinishMoveObject
  • ccb25ca satellite/metabase: add method for retrieving pending object’s metadata
  • a136a7b web/satellite,admin-ui: fix title bug
  • 944ec3b satellite/admin: add OIDC logout
  • c69a69f satellite/{analytics,console}: send segment event when free user logins with free tiers disabled
  • d078b90 satellite/console: add partner to admin email mapping to config
  • c806545 satellite/console: improve sso security
  • 24d95c3 satellite/metainfo: exclude ObjectVersion from non-all-versions listing
  • e8de531 satellite/metabase: filter expires_at in IteratePendingObjectsByKey
  • ed7a72d satellite/admin: add ‘inspect access’ endpoint
  • 0613b06 satellite/admin/ui: add new modal to inspect access
  • 4f914ce satellite/admin: add endpoint to revoke access
  • 8ebed44 satellite/admin/ui: extended access inspector with revoke functionality
  • c7dd510 satellite/eventing: send ObjectRemoved:Delete event for moved objects
  • 457ed6d web/satellite: fix deletes in the object browser
  • ef436fe satellite/eventing: URL-encode object keys in S3 event notifications
  • f2f6eae satellite/admin: extend access inspect feature to handle public S3 keys
  • f2898d5 satellite/admin: extended reg tokens to accept optional partner field
  • 43ead80 satellite/console: notify partner admin when ‘partnered’ reg token is used
  • 3d600c1 satellite/metabase: update CommitSegment to set checksum
  • 098b574 satellite/metabase: update GetSegment to return checksum
  • 5d1f05c web/satellite: fix linter errors and improve typings for the object browser code
  • 5a80ea4 satellite/metainfo: flag to grant OM license for all
  • 0e4520f satellite/metabase: drop emulator workaround for allow_txn_exclusion
  • bcfa264 satellite,testsuite/playwright-ui: fix UI tests
  • 0f4cbd3 satellite/{accounting,satellitedb}: fix missing zero tally for deleted buckets
  • ca82eb8 satellite/admin: add admin email to audit log
  • 9de45b6 satellite/{console,webhook}: improve webhook handling
  • 58aff5b satellite/metainfo: report response size for some of endpoints
  • b334525 satellite/metainfo: defer usage tracking to capture response size
  • e41386a satellite/admin: add disqualify node functionality
  • 52bdceb satellite/repair/repairer: fix data race in counting
  • b209e64 satellite/audit: fix data race on seed
  • ff59f21 satellite/admin: add undisqualify node functionality
  • a8fc332 satellite/console: add bucket creation via public API
  • 7717c49 satellite/payments: fix idempotency for invoice items generation
  • d73efa2 satellite/admin: Move undisqualify node handler
  • c5f6daa satellite/admin/auditlogger: avoid potential send on closed channel
  • b8c560b satellite/jobq: fix data race in Client.Close
  • 7dd325e satellite/jobq: fix data race in QueueMap.StopAll
  • 730fa87 satellite/console: create restricted access via private API
  • b3b800c web/satellite: update large file message
  • db40e88 satellite/nodeselection: clone input before in-place sort
  • 56a7fd1 satellite/overlay: extract selection caches from overlay service
  • 9149dce satellite/metabase: add backend-agnostic ObjectIterator interface
  • 8dac5d7 private/post,satellite/mailservice: improve emails sent from the satellite

Storagenode

  • 891b52c storagenode/piecestore/usedserials: add cleanup chore
  • 1f41940 storagenode/monitor: Fix allocated_space for hashstore only storagenodes
  • 6508ccb storagenode/hashstore: perform lru i/o outside of lock
  • 0ad49fe storagenode/hashstore: bail out of big pages with random writes
  • ca0ff46 storagenode/hashstore: compaction protection for disk usage
  • 8224d2b storagenode/monitor: include hashstore usage in PreFlightCheck totalUsed
  • 1052b3d storagenode/piecestore: fix a small data race
  • 7b5ff4b storagenode/trust: fix data race in access/modifying things
4 Likes
2
3

Perhaps not related, need more data. It could be usual not finished filewalker, so the node didn’t recognize part of the used space and used a minimum between the actual available space and allocated on start.

4

i think it is related , had no issues before , just came up after upgrading to 1.153.2

image
image340×124 1.44 KB

image
image466×121 2.89 KB

image
image984×493 35.4 KB

other node that runs 1.152.3 works as it suppose to

image
image859×472 35.1 KB

Bug in 1.152.5 - disk size
6

Having the same issue 152.5 did with 153.2 on docker node. It isn’t listening to my allocated parameter of 13TB, only using 7.61TB, no free space for ingress.

7

I would suggest to use the recommended version on https://version.storj.io at the moment.
@Morisey thank you for notifying. How is it happened, that your node is updated to the not recommended version? The updater shouldn’t do this, or there is a bug.

8

well since those are kind of shifty boxes i dont use autoupdater , i start services manually and also doing updates manually , so that part is on me , i periodically taking latests from git and updating not to fall behind , so no blame on your part for that , it just that latest versions came out glichy

9

I see, anyway, thank you for notifying. I shared this issue with the team.

10

v1.154.0-rc still broken

11

Regarding log space preallocation, I wonder how is the performance impact of it? I know that when I used preallocation for torrent downloads, the client froze for some seconds. I wonder if isn’t better to preallocate all the allocated space for logs, because the size is the same for all. It could be left some space, like 10%, for other files. This way, the chore will run only once.

12

I’d like to see some numbers myself. Preallocation makes sense for a random writes, sequential reads scenario. Torrents are a perfect example. But here we have sequential writes. I wouldn’t really see preallocation making any difference, actually I would expect that long-term we should be seeing slightly (probably barely measurable) faster operation without preallocation, akin to the delayed allocation feature of some file systems.