searcher: fix cache size env vars rendering 0 for non-Gi storageSize

charts/sourcegraph/CHANGELOG.md

@@ -10,6 +10,7 @@ Use `**BREAKING**:` to denote a breaking change
 
 - Added opt-in Redis config management through `redisCache.config.enabled` and `redisStore.config.enabled` (both default to `false`). Default deployments retain their image configs, custom mounts, authentication, and memory settings. When enabled, the chart mounts `/etc/redis/redis.conf` and supports `existingConfig`, `maxmemory`, `maxmemoryRatio`, `maxmemoryPolicy`, and `additionalConfig`. See [Enabling Redis config management](README.md#enabling-redis-config-management) before opting in, especially when using custom images or Secret-mounted configs.
 - With Redis config management enabled, the chart sizes `maxmemory` at 75% of `resources.limits.memory` to leave room for overhead; this reduces, but does not eliminate, OOM risk. Ratios must be greater than 0 and less than 1. Auto-sizing is skipped in local development mode or when the limit is absent or unrecognised. At the default 7Gi limit, opting in lowers the effective cap from 6GiB to 5.25GiB: `redis-cache` evicts earlier, but `redis-store` uses `noeviction`, so its write-error ceiling drops too. Set `redisStore.config.maxmemory: 6gb` to keep the old ceiling if the pod has sufficient memory.
+- `searcher.storageSize` now accepts any Kubernetes quantity (e.g. `150G`, `1Ti`, `1.5Ti`, `500000M`) when deriving `SEARCHER_CACHE_SIZE_MB` and `SYMBOLS_CACHE_SIZE_MB`. Previously anything other than `Gi` silently rendered both as `0`, which disables cache eviction and lets the cache volume fill up. Unparseable values, or values below `3M`, now fail the render unless `searcher.autoCacheSize` is enabled. The derived size now matches searcher's own auto-sizing arithmetic (45% of the volume in decimal megabytes), so existing `Gi` deployments see a slightly higher value on upgrade, e.g. the default `26Gi` renders `12562` instead of `11700`.
 - Added a `network-policy` example, which limits Executor and Executor job pods to the frontend API
 - Corrected the external object storage examples to configure the shared store for frontend, worker, precise code intel, syntactic code intel, gitserver, and searcher, including credentials or workload service accounts as required.
 - Removed the unused application ports from the precise and syntactic code intel worker Deployments and Services; health checks and Prometheus metrics continue to use the debug server on port 6060.

charts/sourcegraph/README.md

@@ -419,7 +419,7 @@ In addition to the documented values, all services also support the following va
 | searcher.serviceAccount.create | bool | `false` | Enable creation of ServiceAccount for `searcher` |
 | searcher.serviceAccount.name | string | `""` | Name of the ServiceAccount to be created or an existing ServiceAccount |
 | searcher.storageAnnotations | object | `{}` | Optional annotations to add to the `searcher` PVC |
-| searcher.storageSize | string | `"26Gi"` | Size of the PVC for searcher pods to store cache data |
+| searcher.storageSize | string | `"26Gi"` | Size of the PVC for searcher pods to store cache data. Unless `autoCacheSize` is enabled, this must be a Kubernetes quantity (e.g. `150G`, `100Gi`, `1Ti`) of at least `3M`, so the chart can derive `SEARCHER_CACHE_SIZE_MB` and `SYMBOLS_CACHE_SIZE_MB` from it |
 | searcher.storageSubPath | string | `""` | Optional subPath for the `searcher` primary data volume mount |
 | sgTestConnection | object | `{"enabled":true}` | Enable the busybox connection test after deployment |
 | sourcegraph.affinity | object | `{}` | Global Affinity, learn more from the [Kubernetes documentation](https://kubernetes.io/docs/concepts/scheduling-eviction/assign-pod-node/#affinity-and-anti-affinity) |

charts/sourcegraph/templates/_helpers.tpl

@@ -313,6 +313,24 @@ checksum/auth: {{ toJson $checksum | sha256sum }}
 checksum/redis: {{ toJson $checksum | sha256sum }}
 {{- end -}}
 
+{{/*
+Convert a Kubernetes resource quantity (e.g. 7Gi, 1.5G, 500M, 1Ti, 6442450944)
+to a whole number of bytes, rounded down. Renders an empty string when the
+value is not a recognised quantity, so callers can decide whether to fall back
+or fail.
+Usage: include "sourcegraph.quantityBytes" .Values.searcher.storageSize
+*/}}
+{{- define "sourcegraph.quantityBytes" -}}
+{{- $quantity := . | toString -}}
+{{- $number := regexReplaceAll "^([0-9]+(\\.[0-9]+)?).*$" $quantity "${1}" -}}
+{{- $suffix := regexReplaceAll "^[0-9]+(\\.[0-9]+)?" $quantity "" -}}
+{{- /* Kubernetes quantity suffixes: binary (1024^n) and decimal (1000^n). */ -}}
+{{- $units := dict "" 1.0 "k" 1e3 "M" 1e6 "G" 1e9 "T" 1e12 "P" 1e15 "E" 1e18 "Ki" 1024.0 "Mi" 1048576.0 "Gi" 1073741824.0 "Ti" 1099511627776.0 "Pi" 1125899906842624.0 "Ei" 1152921504606846976.0 -}}
+{{- if and (regexMatch "^[0-9]+(\\.[0-9]+)?$" $number) (hasKey $units $suffix) -}}
+{{- printf "%d" (int64 (floor (mulf (float64 $number) (index $units $suffix)))) -}}
+{{- end -}}
+{{- end -}}
+
 {{/*
 Resolve the redis `maxmemory` directive for a service.
 Usage: include "sourcegraph.redis.maxmemory" (list . "redisCache")
@@ -333,20 +351,16 @@ Resolution order:
 {{- if $config.maxmemory -}}
 {{- $config.maxmemory -}}
 {{- else if not $top.Values.sourcegraph.localDevMode -}}
-{{- $limit := dig "resources" "limits" "memory" "" $values | toString -}}
-{{- $number := regexReplaceAll "^([0-9]+(\\.[0-9]+)?).*$" $limit "${1}" -}}
-{{- $suffix := regexReplaceAll "^[0-9]+(\\.[0-9]+)?" $limit "" -}}
-{{- /* Kubernetes quantity suffixes: binary (1024^n) and decimal (1000^n). */ -}}
-{{- $units := dict "" 1.0 "k" 1e3 "M" 1e6 "G" 1e9 "T" 1e12 "P" 1e15 "E" 1e18 "Ki" 1024.0 "Mi" 1048576.0 "Gi" 1073741824.0 "Ti" 1099511627776.0 "Pi" 1125899906842624.0 "Ei" 1152921504606846976.0 -}}
-{{- if and (regexMatch "^[0-9]+(\\.[0-9]+)?$" $number) (hasKey $units $suffix) -}}
+{{- $limit := include "sourcegraph.quantityBytes" (dig "resources" "limits" "memory" "" $values) -}}
+{{- if $limit -}}
 {{- $ratio := 0.75 -}}
 {{- if hasKey $config "maxmemoryRatio" -}}
 {{- $ratio = float64 $config.maxmemoryRatio -}}
 {{- end -}}
 {{- if not (and (gt $ratio 0.0) (lt $ratio 1.0)) -}}
 {{- fail (printf "%s.config.maxmemoryRatio must be greater than 0 and less than 1" $service) -}}
 {{- end -}}
-{{- $bytes := floor (mulf (float64 $number) (index $units $suffix) $ratio) -}}
+{{- $bytes := floor (mulf (float64 $limit) $ratio) -}}
 {{- if gt $bytes 0.0 -}}
 {{- printf "%d" (int64 $bytes) -}}
 {{- end -}}

charts/sourcegraph/templates/_helpers/_searcherCacheSize.tpl

@@ -0,0 +1,22 @@
+{{/*
+45% of searcher.storageSize in whole (decimal) megabytes, for
+SEARCHER_CACHE_SIZE_MB and SYMBOLS_CACHE_SIZE_MB. Mirrors the arithmetic
+searcher uses when it auto-sizes from the volume: floor(bytes / 1e6) * 45 / 100.
+Accepts any Kubernetes quantity (e.g. 150G, 100Gi, 1.5Ti, 500000M).
+
+Anything unparseable, or anything so small the result would be 0, fails the
+render: searcher treats an explicit 0 as "never evict" (unlike an unset
+variable, which auto-sizes), so the cache volume would fill up.
+*/}}
+{{- define "sourcegraph.searcher.cacheSizeMB" -}}
+{{- $size := .Values.searcher.storageSize | default "26Gi" | toString -}}
+{{- $bytes := include "sourcegraph.quantityBytes" $size -}}
+{{- if not $bytes -}}
+{{- fail (printf "searcher.storageSize must be a Kubernetes quantity such as 150G, 100Gi or 1Ti (got %q), or set searcher.autoCacheSize=true" $size) -}}
+{{- end -}}
+{{- $megabytes := div (mul (div (int64 $bytes) 1000000) 45) 100 -}}
+{{- if le $megabytes 0 -}}
+{{- fail (printf "searcher.storageSize %s is too small to derive a non-zero cache size; use at least 3M, or set searcher.autoCacheSize=true" $size) -}}
+{{- end -}}
+{{- $megabytes -}}
+{{- end -}}

charts/sourcegraph/templates/searcher/searcher.StatefulSet.yaml

@@ -62,24 +62,11 @@ spec:
         # sizes its cache to ~45% of the live cache volume, tracking PVC
         # expansion automatically. SEARCHER_CACHE_SIZE_MB / SYMBOLS_CACHE_SIZE_MB
         # are deprecated in favor of this auto-detection.
+        # Both caches share the PVC below and each gets ~45% of it.
         - name: SEARCHER_CACHE_SIZE_MB
-          # Set the cache size to ~45% of the PVC size
-          {{- if .Values.searcher.storageSize }}
-          value: {{ trimSuffix "Gi" .Values.searcher.storageSize | mul 450 | quote }}
-          {{- else }}
-          # This value is ~45% of the default value for
-          # storageSize in the VolumeClaimTemplate below
-          value: "11700"
-          {{- end }}
+          value: {{ include "sourcegraph.searcher.cacheSizeMB" . | quote }}
         - name: SYMBOLS_CACHE_SIZE_MB
-          # Set the cache size to ~45% of the PVC size
-          {{- if .Values.searcher.storageSize }}
-          value: {{ trimSuffix "Gi" .Values.searcher.storageSize | mul 450 | quote }}
-          {{- else }}
-          # This value is ~45% of the default value for
-          # storageSize in the VolumeClaimTemplate below
-          value: "11700"
-          {{- end }}
+          value: {{ include "sourcegraph.searcher.cacheSizeMB" . | quote }}
         {{- end }}
         - name: POD_NAME
           valueFrom:

charts/sourcegraph/tests/searcherCacheSize_test.yaml

@@ -0,0 +1,124 @@
+suite: searcher cache size
+templates:
+- searcher/searcher.StatefulSet.yaml
+tests:
+- it: should size both caches to 45% of the default storageSize
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "12562"
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SYMBOLS_CACHE_SIZE_MB
+        value: "12562"
+- it: should accept storageSize in Gi
+  set:
+    searcher:
+      storageSize: 100Gi
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "48318"
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SYMBOLS_CACHE_SIZE_MB
+        value: "48318"
+- it: should accept storageSize in G
+  set:
+    searcher:
+      storageSize: 150G
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "67500"
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SYMBOLS_CACHE_SIZE_MB
+        value: "67500"
+  - equal:
+      path: spec.volumeClaimTemplates[0].spec.resources.requests.storage
+      value: 150G
+- it: should accept storageSize in M
+  set:
+    searcher:
+      storageSize: 500000M
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "225000"
+- it: should accept storageSize in Ti
+  set:
+    searcher:
+      storageSize: 1Ti
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "494779"
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SYMBOLS_CACHE_SIZE_MB
+        value: "494779"
+- it: should accept a fractional storageSize
+  set:
+    searcher:
+      storageSize: 1.5G
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "675"
+- it: should accept a suffixless byte count and round down to whole megabytes
+  set:
+    searcher:
+      storageSize: "6442450944"
+  asserts:
+  - contains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+        value: "2898"
+- it: should fail on an unrecognised quantity instead of disabling eviction
+  set:
+    searcher:
+      storageSize: 1Tb
+  asserts:
+  - failedTemplate:
+      errorMessage: searcher.storageSize must be a Kubernetes quantity such as 150G, 100Gi or 1Ti (got "1Tb"), or set searcher.autoCacheSize=true
+- it: should fail when the derived cache size would be 0 instead of disabling eviction
+  set:
+    searcher:
+      storageSize: 2M
+  asserts:
+  - failedTemplate:
+      errorMessage: searcher.storageSize 2M is too small to derive a non-zero cache size; use at least 3M, or set searcher.autoCacheSize=true
+- it: should not need a parseable storageSize when autoCacheSize is enabled
+  set:
+    searcher:
+      autoCacheSize: true
+      storageSize: 1Tb
+  asserts:
+  - notContains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SEARCHER_CACHE_SIZE_MB
+      any: true
+  - notContains:
+      path: spec.template.spec.containers[0].env
+      content:
+        name: SYMBOLS_CACHE_SIZE_MB
+      any: true

charts/sourcegraph/values.yaml

@@ -1282,7 +1282,9 @@ searcher:
   # This tracks PVC expansion automatically, whereas the hardcoded env vars are
   # frozen to the initial `storageSize` and do not follow later disk growth.
   autoCacheSize: false
-  # -- Size of the PVC for searcher pods to store cache data
+  # -- Size of the PVC for searcher pods to store cache data. Unless `autoCacheSize`
+  # is enabled, this must be a Kubernetes quantity (e.g. `150G`, `100Gi`, `1Ti`) of at
+  # least `3M`, so the chart can derive `SEARCHER_CACHE_SIZE_MB` and `SYMBOLS_CACHE_SIZE_MB` from it
   storageSize: 26Gi
   # -- Optional subPath for the `searcher` primary data volume mount
   storageSubPath: ""