Log Recognition / Family call fails with 503 error

Wellbore Log Recognition / Family call fails with 503 error intermittently.

1. Need to understand and resolve the error.
2. Additionally, it is critical to understand if during this failure there is any data loss or lack of data mapping.
3. Ask to enhance documentation from the domain side for how the call works today

added Azure label

@mgherzog

Following the ask items above,

1. Could you provide more details on,

   • Is the behavior reproducible in the Azure environment used for the GitLab CI/CD?
   • Is it reproducible only in Azure or also other CSPs environments?
   • Is this behavior dependent on the payload used? Could you share the curl used ?
   • How often does it happen?
   • Could you share the test set/utility to reproduce it, if applicable?
   • Could you share system and service monitoring metrics (CPU, memory, pod count, number of requests, ...)?
   • Which are the resource/request limits, and HPA/KEDA configuration used?

2. No data is persisted or updated when using log recognition GET API

   • No data loss or data corruption involved

3. Could you have a look at the current documentation at https://community.opengroup.org/osdu/platform/domain-data-mgmt-services/wellbore/wellbore-domain-services/-/wikis/Log-recognition and provide more insights on which FAQ/topics are missing or need more clarification?

CC @sumrazafar @Dmitriy_Rudko

Hi @kin.ng

Adding @dacarpen for 1.

• We do not have visibility into other CSP environments
• The command is not precisely understood yet - we are responding to error messages that we've observed on behalf of users.
• Frequency is not predictable but it is reproducible - looks data dependent

2. This is reassuring to know there is no data loss. However, if there is an issue with the service, will the customer have to re-do the call or for example, is the failure a response to no data being present that could be leverage by the family API and therefore if the customer reran, they would face the same issue.
3. Pre-requesites and where to find them in the data. Additionally, it would be great to provide guidance on troubleshooting for data specific concerns and how to understand errors from the service.

Here is the curl used:

curl --location '<url>/api/os-wellbore-ddms/log-recognition/family' \
--header 'data-partition-id: <data partition>' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer <token>' \
--data '{
  "label": "GRD",
  "log_unit": "GAPI",
  "description": "LDTD Gamma Ray"
}'

The failure does not depend on the payload used. This was done with no HPA enabled and no rate limiting.

@dacarpen Could you share more information on

• how to reproduce the issue
• deployment configuration
• monitoring metrics around the time the issue is observed

See comment above

So far it looks like CSP specific, and infra related, rather than an underlying issue with WDMS as a service.