[receiver/hostmetricsreceiver] Align hugepages instrument types (#47217)

<!--Ex. Fixing a bug - Describe the bug and how this fixes the issue.
Ex. Adding a feature - Explain what this achieves.-->
#### Description
`system.memory.linux.hugepages.page_size`,
`system.memory.linux.hugepages.reserved`, and
`system.memory.linux.hugepages.surplus` should be emitted as
updowncounters rather than gauges.

Context:
open-telemetry/semantic-conventions#3516 (comment)

<!-- Issue number (e.g. #1234) or full URL to issue, if applicable. -->

Signed-off-by: Shaanveer Singh <shaanver.singh@gmail.com>
Co-authored-by: Braydon Kains <93549768+braydonk@users.noreply.github.com>

Author: ShaanveerS

.chloggen/hostmetrics-hugepages-instrument-types.yaml

@@ -0,0 +1,11 @@
+change_type: bug_fix
+
+component: receiver/hostmetrics
+
+note: Align HugePages metric instrument types with the semantic conventions by emitting page_size, reserved, and surplus as non-monotonic sums instead of gauges.
+
+issues: [42650]
+
+subtext:
+
+change_logs: [user]

receiver/hostmetricsreceiver/internal/scraper/memoryscraper/documentation.md

@@ -72,25 +72,25 @@ Total number of hugepages available.
 
 System hugepage size in bytes.
 
-| Unit | Metric Type | Value Type | Stability |
-| ---- | ----------- | ---------- | --------- |
-| By | Gauge | Int | Development |
+| Unit | Metric Type | Value Type | Aggregation Temporality | Monotonic | Stability |
+| ---- | ----------- | ---------- | ----------------------- | --------- | --------- |
+| By | Sum | Int | Cumulative | false | Development |
 
 ### system.memory.linux.hugepages.reserved
 
 Number of reserved hugepages (hugepages for which a commitment to allocate has been made, but no allocation has yet been made). This is reported as a separate metric rather than a usage state because reserved pages are already counted in free pages - they represent a subset of free pages that cannot be used for non-reserved allocations.
 
-| Unit | Metric Type | Value Type | Stability |
-| ---- | ----------- | ---------- | --------- |
-| {page} | Gauge | Int | Development |
+| Unit | Metric Type | Value Type | Aggregation Temporality | Monotonic | Stability |
+| ---- | ----------- | ---------- | ----------------------- | --------- | --------- |
+| {page} | Sum | Int | Cumulative | false | Development |
 
 ### system.memory.linux.hugepages.surplus
 
 Number of surplus hugepages (overcommitted hugepages beyond the persistent pool). This is reported as a separate metric rather than a usage state because surplus pages can be in either used or free state, and including them would break the semantic convention that usage states must sum to the limit.
 
-| Unit | Metric Type | Value Type | Stability |
-| ---- | ----------- | ---------- | --------- |
-| {page} | Gauge | Int | Development |
+| Unit | Metric Type | Value Type | Aggregation Temporality | Monotonic | Stability |
+| ---- | ----------- | ---------- | ----------------------- | --------- | --------- |
+| {page} | Sum | Int | Cumulative | false | Development |
 
 ### system.memory.linux.hugepages.usage
 

receiver/hostmetricsreceiver/internal/scraper/memoryscraper/internal/metadata/generated_metrics.go

@@ -70,7 +70,7 @@ metrics:
     description: System hugepage size in bytes.
     unit: By
     stability: development
-    gauge:
+    sum:
       value_type: int
       aggregation_temporality: cumulative
       monotonic: false
@@ -80,16 +80,20 @@ metrics:
     description: Number of reserved hugepages (hugepages for which a commitment to allocate has been made, but no allocation has yet been made). This is reported as a separate metric rather than a usage state because reserved pages are already counted in free pages - they represent a subset of free pages that cannot be used for non-reserved allocations.
     unit: "{page}"
     stability: development
-    gauge:
+    sum:
       value_type: int
+      aggregation_temporality: cumulative
+      monotonic: false
 
   system.memory.linux.hugepages.surplus:
     enabled: false
     description: Number of surplus hugepages (overcommitted hugepages beyond the persistent pool). This is reported as a separate metric rather than a usage state because surplus pages can be in either used or free state, and including them would break the semantic convention that usage states must sum to the limit.
     unit: "{page}"
     stability: development
-    gauge:
+    sum:
       value_type: int
+      aggregation_temporality: cumulative
+      monotonic: false
 
   system.memory.linux.hugepages.usage:
     enabled: false