Aleksei Sviridkin
e9411d6b8c
The qdrant and harbor post-delete hooks guarded every delete with `|| echo "WARNING: ..." >&2`, set no `-e`, and ended on an unconditional completion message, so the container exited 0 on every path. Helm read that as a successful hook and the hook-succeeded delete policy removed the Job together with the pod that carried the warning, leaving an uninstall that reclaimed nothing indistinguishable from one that had nothing to reclaim. Count the failures instead and set the exit code at the end, so a failed delete does not skip the deletes that follow it. An object that is already gone stays a success: every delete keeps --ignore-not-found, which exits 0 in that case. This is the shape packages/apps/mariadb already uses for the same problem. Both hooks also consumed the output of `kubectl get` directly -- harbor by iterating it, qdrant by piping it into grep, whose status the pipeline then reports. A listing that failed therefore arrived as an empty list and the sweep was skipped rather than reported. Each listing is now captured before it is used, so a listing that failed and one that returned nothing are different events. A hook that exits non-zero keeps its Job and pods, which is the point: they hold the log. Give the Job one retry rather than the Kubernetes default of six or none at all. Every delete is idempotent, so a second attempt re-issues an API call rather than re-running any logic, and one short backoff is what separates a transient apiserver error from a real one -- these failures become visible for the first time here, and a signal that fires on noise stops being read, including on the day it is right. Six is the far end of the same mistake: ten minutes of backoff across seven pods outlasts the hook timeout, and Helm reports that timeout instead of the hook's own error. The qdrant Chainsaw suite explained its post-delete pod collector by quoting the guard string removed here and concluding the pod is always collected away. That is now backwards: a failed cleanup is exactly when the pod survives and the collector is the evidence. Rewrite the note, and the sentence in the harbor suite pointing at it. hack/app-cleanup-hooks_test.bats renders both scripts and runs them against a fake kubectl. There are eight failure paths across the two hooks -- six in harbor, two in qdrant -- and each has a case that goes red when that one path alone is reverted to a swallowed warning. Message assertions are anchored on the hook's own `ERROR: ` prefix, because the fake kubectl names the object in its own diagnostic too and an unanchored match is satisfied by the fixture. The helm-unittest suites pin the retry limit and the hook delete policy: adding hook-failed to that annotation hands the Job back to Helm to delete on failure and destroys the evidence this change exists to keep, and until these asserts that reversal left both suites green. Assisted-By: Claude <noreply@anthropic.com> Signed-off-by: Aleksei Sviridkin <f@lex.la>
Managed Qdrant Service
Qdrant is a high-performance vector database and similarity search engine designed for AI and machine learning applications. It provides efficient storage and retrieval of high-dimensional vectors with advanced filtering capabilities, making it ideal for recommendation systems, semantic search, and RAG (Retrieval-Augmented Generation) applications.
Deployment Details
Service deploys Qdrant as a StatefulSet with automatic cluster mode when multiple replicas are configured.
storageClassis annotated as immutable in the chart schema — seedocs/storage-immutability.mdfor the contract and which consumers enforce it.
Parameters
Common parameters
| Name | Description | Type | Value |
|---|---|---|---|
replicas |
Number of Qdrant replicas. Cluster mode is automatically enabled when replicas > 1. | int |
1 |
resources |
Explicit CPU and memory configuration for each Qdrant replica. When omitted, the preset defined in resourcesPreset is applied. |
object |
{} |
resources.cpu |
CPU available to each replica. | quantity |
"" |
resources.memory |
Memory (RAM) available to each replica. | quantity |
"" |
resourcesPreset |
Default sizing preset used when resources is omitted. |
string |
t1.small |
size |
Persistent Volume Claim size available for vector data storage. | quantity |
10Gi |
storageClass |
StorageClass used to store the data. | string |
"" |
external |
Enable external access from outside the cluster. | bool |
false |
TLS parameters
| Name | Description | Type | Value |
|---|---|---|---|
tls |
TLS configuration. When omitted, the effective TLS state follows the value of external. |
object |
{} |
tls.enabled |
Enable TLS. When unset, inherits the value of external (TLS is on when external access is enabled). Set explicitly to true or false to override. |
*bool |
null |
Parameter examples and reference
resources and resourcesPreset
resources sets explicit CPU and memory configurations for each replica.
When left empty, the preset defined in resourcesPreset is applied.
resources:
cpu: 4000m
memory: 4Gi
resourcesPreset sets named CPU and memory configurations for each replica.
This setting is ignored if the corresponding resources value is set.
Presets follow a cloud-style <series>.<size> naming convention. Five series cover the full CPU-to-memory ratio range (t1 1:0.5, c1 1:1, s1 1:2, u1 1:4, m1 1:8) and each series ships eight sizes (nano through 4xlarge). The legacy flat names (nano, micro, small, medium, large, xlarge, 2xlarge) remain accepted as deprecated aliases of their 1:1 instance-type equivalents.
See docs/operations/resource-presets.md for the full size matrix and the legacy-to-instance-type mapping.