Apheris Hub - Known Issues🔗
Below, we list known issues with Apheris Hub. If available, we list a workaround to help troubleshoot the issue.
This list will be constantly evolving with input from our different users and with the evolution of our product in future releases.
Known Issues with Apheris Hub 0.3.1🔗
OpenFold3🔗
| Issue | Description | Additional information |
|---|---|---|
| Unexpected Behavior on some Batch Requests | Due to an unresolved MSA cache bug in the OpenFold3 code base, batch requests (i.e. requests with more than one query) with more than a single unique protein sequence and counterpart MSA may cause unexpected behavior. | Please submit individual requests with a single query for these complexes. |
| MSA Files Cannot Be Used with OpenFold3 | An issue prevents the reliable usage of MSA files with the current OpenFold3 model. | Do not use MSA files with the OpenFold3 model version. The public MSA server version is available, but you should not use MSA files with it. |
Inference and Processing🔗
| Issue | Description | Additional information |
|---|---|---|
| ColabFold Server Rate Limiting | The public ColabFold server applies rate limiting, which can significantly slow down processing of even simple queries. | If you encounter slower than anticipated query times it is likely due to using the public ColabFold server. Please wait a few minutes and try again or use MSAs from file if the model supports it. |
| Request Status Tracking | Inference requests can get stuck in a "running" state if the corresponding model is stopped or uninstalled during processing. | After an instance restart, the request is transitioned to done, but the results are not displayed. |
| Results Polling | When going to the results page before a prediction completed, you might have to refresh the page for the results to load. | |
| Delayed Execution on Shared GPU | When we have a single GPU being shared between two running model versions, parallel jobs can be started on each of those models. However, because they share the same GPU, the execution can be blocked and delayed. | Additionally, the UI shows the delayed requests as "Running" instead of "Pending", as they are waiting for the GPU to become available. |
| Error Log Display | Error logs are shown even for successful inference runs. However, they don't always contain errors per se but just the stderr output. |
|
| Job Duration Metrics | Job duration doesn't distinguish between waiting time and actual running time, which can be misleading when multiple jobs are queued. | |
| Accepted jobs can transition directly to done leading to issues in the queue | Jobs that complete faster than the current polling interval (1 second) may transition directly from Accepted to Done, bypassing intermediate states. | This can cause inconsistencies or potential issues with the job queue handling. |
| Pending or accepted jobs do not time out | Pending or accepted requests currently do not time out as expected. Requests that remain in a pending state indefinitely can lead to queue issues. | |
| Uninstalling a model should set all jobs that are accepted/running to failed and cancel all ongoing tasks | When uninstalling a model, jobs that are in Accepted or Running states should be marked as Failed, and any ongoing tasks should be canceled. | Currently, these jobs may remain in the queue or continue running even after the model is uninstalled, leading to inconsistencies. |
Security🔗
| Issue | Description | Additional information |
|---|---|---|
| Container User Privileges | Docker containers are currently running with root user privileges, which is not a security best practice. |
User Interface🔗
| Issue | Description | Additional information |
|---|---|---|
| PAE/PDE Plot Selection | When trying to select areas of the PAE/PDE plot, the selection box behaves erratically and lower areas cannot be selected. It is not possible to highlight the edges of the plot. | |
| UI behavior when the API key is wrong | The Hub displays progress bars and uninstall button for failed model installations with a wrong API key. | |
| Erratic behavior when cancelling a fast installation | When cancelling a fast installation, hitting cancel leads to an erratic behavior. | The model has the uninstalled state and an associated error, although the installation was cancelled successfully as per the success message. |
| Inconsistency when searching by job ID | Searching for a unique job ID doesn't consistently return a single result. | |
| Missing pLDDT coloring for OpenFold3 | pLDDT coloring for OpenFold3 is not yet available, leading to inconsistent visualization when compared to Boltz-2 models. | |
| Incorrect SMILES sequence on multi-protein ligand example | When using the multi-protein ligand example, the SMILES sequence shown on the results view is the same on both tabs. |
Form and Input Handling🔗
| Issue | Description | Additional information |
|---|---|---|
| Queries editor validation errors | Queries editor shows duplicate validation errors when upload MSA toggle is active. |
Installation and Configuration🔗
| Issue | Description | Additional information |
|---|---|---|
| Port Allocation | The installation script uses a fixed port (8080) which may already be in use on some systems. | As a workaround, ensure port 8080 is available before installation or modify the port in the installation script. |
| CloudFormation Upgrade | Using the CloudFormation template, upgrading an existing stack to a newer version fails. | |
| CloudFormation CIDR Range Parameter Limitations | The CloudFormation template doesn't allow specifying multiple CIDR ranges in the AllowIngressFromCidr parameter, limiting access control options. |
|
| CloudFormation Deployment Failures | CloudFormation deployments occasionally fail without clear error messages. | |
| CloudFormation Roll Back Limitation | If the Hub doesn't start, the CloudFormation stack is not rolled back. | |
| Feedback on Wrong Configuration or Systems Faults | On installation, if a configuration item is wrong (e.g. the API key) or a system fault occurs (e.g. no space left on device), there is no feedback to the user on UI. |
Inspect the Hub logs (see View Docker Logs to troubleshoot the issue. |
| CloudFormation stack name limitation | Hub CloudFormation stack fails when name exceeds 64 characters due to IAM role naming constraints. | |
| Missing tag uniqueness validation | In the data.yaml configuration, there is no validation on tag uniqueness. |