Moved backend to new subnet - front-end unhappy

Hi,

Our backend servers have moved to a different facility and are now on a new subnet. The data is the same, the machines are the same. But the IP addresses have changed.

We have a very small labeling project running on the heartexlabs/label-studio:latest docker hub image and backed by a Minio instance that provides an s3 compatible object store.

To accommodate the IP change, I went into the Settings->Cloud Storages page and updated the IP. The “check connection” works.

After changing the IPs, I reran the “sync storage” process.

My use of the system is mainly through the API to pull images and annotations out of the LabelStudio projects. It all works - I can retrieve annotations and grab images through the path LabelStudio returns in the task’s JSON under data.image.

The “Label all Tasks” UI seems to work just fine.

However, the project UI that normally lists all tasks (/projects/2/data?tab=1) is empty. In the browser’s JavaScript console there are a large number of the following errors:

Error: [mobx-state-tree] You are trying to read or write to an object that is no longer part of a state tree. (Object type: 'AnnotationStore', Path upon death: '/annotationStore', Subpath: 'selected', Action: ''). Either detach nodes first, or don't use objects after removing / replacing them in the tree.
    at De (mobx-state-tree.module.js:3427:18)
    at t.assertAlive (mobx-state-tree.module.js:1510:21)
    at t.unbox (mobx-state-tree.module.js:1580:14)
    at t.dehanceValue (mobx.module.js:1040:25)
    at t.get (mobx.module.js:1092:21)
    at e.read (mobx.module.js:4136:37)
    at Object.get [as selected] (mobx.module.js:4377:36)
    at DraftGuard.js:33:55
    at Object.confirmTransitionTo (history.js:131:51)
    at push (history.js:350:23)
    at navigate (react-router-dom.js:189:83)
    at onClick (react-router-dom.js:145:11)
    at Object.He (react-dom.production.min.js:52:317)
    at Ze (react-dom.production.min.js:52:471)
    at react-dom.production.min.js:53:35
    at Tr (react-dom.production.min.js:100:68)
    at jr (react-dom.production.min.js:101:380)
    at react-dom.production.min.js:113:65
    at De (react-dom.production.min.js:292:189)
    at react-dom.production.min.js:50:57
    at Pr (react-dom.production.min.js:105:469)
    at Qt (react-dom.production.min.js:75:265)
    at Xt (react-dom.production.min.js:74:124)
    at t.unstable_runWithPriority (scheduler.production.min.js:18:343)
    at Vi (react-dom.production.min.js:122:325)
    at Ae (react-dom.production.min.js:292:48)
    at Yt (react-dom.production.min.js:73:352)
    at HTMLDivElement.i (helpers.js:74:23)

How can this be resolved?

Versions:
Label Studio via docker heartexlabs/label-studio:latest:

{
  "release": "1.12.0",
  "label-studio-os-package": {
    "version": "1.12.0",
    "short_version": "1.12",
    "latest_version_from_pypi": "1.13.1",
    "latest_version_upload_time": "2024-08-20T17:38:07",
    "current_version_is_outdated": true
  },

  "label-studio-os-backend": {
    "message": "fix: LEAP-947: Check task when we get a response from ML (#5742)",
    "commit": "b2fb5819f214d06f9300bd6c3a7853c78bdd176d",
    "date": "2024/04/17 15:26:37",
    "branch": "",
    "version": "1.12.0+0.gb2fb581"
  },

  "label-studio-converter": {
    "version": "0.0.58"
  }
}

Thanks!

Please switch to LS 1.13.1, it’s a chromium browser issue after v128.

Thank you! The timing for an auto Chrome update was horrible. Updating Label Studio did the trick.