Known Issues¶
This article lists the known issues and limitations of Enterprise Analytics Platform products.
MI Hub¶
| Issue Description | Discovered On | Resolved On |
|---|---|---|
| Due to the resource and permission restrictions of the underlying image repository, the amount of registered models and model versions is limited. Currently, at most 30 models can be registered, and at most 50 versions can be staged for a model. After reaching the limit, users need to contact the operation team to remove invalid models from the image repository. After release resources, users can register new models and stage new model versions. | 2020/12/30 | |
| In model testing, the size of the test file must not exceed 3.2Kb by default, and the resource for testing is 1vCore 1GB by default. If the testing file size or resource exceeds the limit, the model testing will fail. | 2020/12/30 | |
| In model indicator monitoring, at most 10 customized monitoring reports can be created. Exceeding this limit will result in a significant delay in loading the model. | 2020/12/30 | |
| When saving a model in MLflow, the path must not include “-“. Otherwise, an error will be reported when staging the model version in MI Hub. | 2020/12/30 | |
| The current EAP version does not support deploying models on Edge or registering Edge computing devices. | 2020/12/30 | |
| The current EAP version does not support coprocessors. | 2020/12/30 |
MI Lab¶
| Issue Description | Discovered On | Resolved On |
|---|---|---|
| If the Data IDE account of the OU is not a standard account (not in the format of data_ouid), and Mode Hadoop PVC is selected when creating a Notebook, the Notebook might fail to start due to missing keytab file. | 2020/12/30 | |
| Package installation is not supported on private cloud instances. | 2020/12/30 | |
| Git exception is reported when the Notebook instance uses Azure storage. | 2020/12/30 |
MI Pipelines¶
| Issue Description | Discovered On | Resolved On |
|---|---|---|
Experiment list display runs timeout when a huge amount of pipeline instances (for example, more than 10,000) are running. In this case, run the script runs_cleanup.ipynb to clean running instances. |
2020/12/30 | |
| When the ParallelFor operator is the last operator in a pipeline and an OnExit operator is used, an internal exception error will be reported when viewing the DAG graph of the pipeline. | 2020/12/30 | |
| If a system defined operator is moved to the directory of custom operators, it cannot be restored to the original directory. | 2020/12/30 | |
| Only 1 Recursion operator can be used in a pipeline. | 2020/12/30 |
Resource Configuration¶
| Issue Description | Discovered On | Resolved On |
|---|---|---|
| Due to security control, only administrators can create data source connections, and only the creator can test, modify, and delete data source connections. | 2020/12/30 | |
| If GitLab data source connections have project name with space, an error about empty pointer will be reported. | 2020/12/30 | |
| Users cannot create data source connections with SSO accounts. | 2020/12/30 | |
| GitHub data source connections do not support private repositories. | 2020/12/30 |
Other¶
| Issue Description | Discovered On | Resolved On |
|---|---|---|
| When upgrading the EAP deployment, the original permission configuration might be lost. After the upgrade, you need to check the permission configuration. If the configuration is lost, contact the administrators to add needed permission. | 2020/12/30 |