netdata · github-actions · Aug 29, 2026
diff --git a/docs/Collecting Metrics/Ceph.mdx b/docs/Collecting Metrics/Ceph.mdx
@@ -14,7 +14,11 @@ slug: "/collecting-metrics/ceph"
 
 # Monitor Ceph
 
-Netdata gives you a complete operational view of Ceph by collecting four complementary telemetry surfaces: the MGR Prometheus module, official `ceph-exporter`, the NVMe-oF gateway exporter, and the Ceph Dashboard API. Deploy one Agent close to each Ceph node to monitor cluster state, daemon health, host resources, RGW traffic, and local hardware together.
+Netdata gives you a complete operational view of Ceph by collecting four complementary telemetry surfaces—the MGR
+Prometheus module, official `ceph-exporter`, the NVMe-oF gateway exporter, and the Ceph Dashboard API—and by running
+authenticated S3 lifecycle checks from the client vantages that depend on object storage.
+Deploy one Agent close to each Ceph node to monitor cluster state, daemon health, host resources, RGW traffic,
+client-visible S3 correctness, and local hardware.
 
 ## What you can monitor
 
@@ -23,6 +27,8 @@ Netdata gives you a complete operational view of Ceph by collecting four complem
 - CephFS/MDS, RBD, RBD Mirror, SMB, and client I/O telemetry exposed by your Ceph release.
 - Host-local daemon performance from official `ceph-exporter`.
 - RGW requests, Lua execution, notifications, queues, retries, and access logs.
+- Authenticated S3 write, read, list, delete, payload-integrity, cleanup, and latency results from selected vantages.
+- Directional multisite S3 replication, payload integrity, recovery-point objective, and delete-propagation results.
 - RGW endpoint availability and TLS certificate health.
 - NVMe-oF gateway, block-device, host, subsystem, and namespace telemetry from supported exporters.
 - Node hardware health, cooling, power, memory, processors, storage, and temperature reporting on Tentacle.
@@ -63,6 +69,8 @@ Use this map to identify the Netdata surface that owns the operational question
 | RBD mirroring | Local/remote snapshot timestamp synchronization | MGR Prometheus | Identify mirrored images that have diverged |
 | RGW service health | Notifications, Lua execution, queue pressure, retries, aborted requests | MGR Prometheus | Inspect aggregate gateway behavior |
 | RGW request outcomes | Status classes, bytes, clients, and request duration | `web_log` | Analyze complete RGW access logs |
+| Authenticated S3 correctness | PUT, GET integrity, LIST, DELETE, cleanup, and latency | `s3check` | Verify client-visible object operations |
+| Multisite S3 replication | Directional payload integrity, visibility lag, RPO, and delete propagation | `s3check` | Verify the replication paths that client applications depend on |
 | RGW endpoint reachability | Unauthenticated HTTP liveness | `httpcheck` | Verify the selected endpoint is reachable |
 | RGW certificates | Certificate expiration and revocation | `x509check` | Track certificate lifecycle independently of RGW traffic |
 | Node health | CPU, memory, disk I/O, filesystems, network interfaces, and processes | Standard Netdata collectors | Continue using normal Agent monitoring for each Ceph node |
@@ -100,6 +108,8 @@ Run one Agent on each Ceph node. Each Agent monitors the Ceph services and host
 | `ceph-exporter` | One job on each node whose daemon telemetry you need | Host-local daemon performance and daemon inventory |
 | NVMe-oF gateway exporter | One job on each gateway endpoint | Gateway-local runtime, block-device, host, subsystem, and namespace telemetry |
 | Ceph Dashboard API | One logical job per Ceph cluster | Dashboard API component integrity and Ceph investigation Functions |
+| Authenticated S3 check | One job for each selected client vantage | Client-visible S3 object lifecycle and latency |
+| Directional multisite S3 check | One explicit source-to-destination job per replication path | Client-visible replication correctness, RPO, and delete propagation |
 | Host collectors | Every Agent | Node disks, filesystems, network interfaces, processes, and logs |
 
 Use one stable job identity for the MGR surface. If the active MGR moves, update DNS or the reverse proxy to the current active endpoint rather than creating one job for every possible MGR. Multiple active MGR jobs for the same cluster create duplicate cluster alert owners.
@@ -128,6 +138,37 @@ Enable the exporter in each Ceph NVMe-oF gateway deployment and collect every ga
 
 Enable the Ceph Dashboard module, secure it with TLS, and create a read-only Dashboard user. Configure one native Ceph collector job per cluster. The Dashboard collector complements the metric endpoints: it owns API component integrity and provides Ceph investigation Functions.
 
+### Authenticated S3 checks
+
+Configure an `s3check` job for every client vantage whose object-storage behavior matters. Each job uses a dedicated
+unversioned bucket and prefix, reconciles that prefix, performs one authenticated PUT, GET, LIST, DELETE, and
+cleanup cycle, verifies the downloaded payload, and removes probe objects after interrupted cycles. Place jobs at each
+site or RGW client path that requires a client-visible correctness signal.
+
+For multisite replication, set `mode: multisite` and configure one explicit source and destination. The source uses the
+job's top-level S3 settings; the destination has its own endpoint, region, bucket, prefix, credentials, addressing, and
+transport settings. Add bounded `source_site` and `destination.site` labels, then create one job for each direction you
+want to verify—for example site-a to site-b and site-b to site-a. Netdata never probes every combination automatically.
+The destination prefix identifies where the replicated probe key is expected. If source and destination prefixes differ,
+the replication policy must map the source route namespace onto the destination prefix.
+
+After a multisite job deletes its exact source and destination probe keys, Netdata keeps the sanitized ownership journal until
+the larger configured replication or delete deadline elapses. It then lists both owner-scoped namespaces, waits one more
+collection interval, and repeats the lists in reverse endpoint order before releasing ownership. This bounded confirmation window
+aligns object cleanup with the replication policy you configured.
+
+Configure endpoint addresses that resolve to distinct S3 services; Netdata rejects literal, default-port, and
+virtual-host aliases for the same bucket, but it does not resolve DNS names to guess whether two services share one
+gateway.
+
+A multisite job writes one small source object, verifies the destination object's SHA-256 digest, measures how long
+client visibility takes, deletes the source, and optionally waits for the destination copy to disappear. It persists a
+sanitized ownership journal across Agent restarts, reconciles both Agent-and-job-owned key namespaces before
+new writes, and removes both objects when a visibility or delete deadline is reached. Set `rpo_threshold_ms`,
+`replication_timeout_ms`, `delete_threshold_ms`, `delete_timeout_ms`, and `verify_delete` to match the replication
+policy. Visibility and delete objectives must be at least one collection interval because Netdata polls each bounded
+phase once per cycle; the two objective alerts are silent until you enable and tune them. Probe keys live in an Agent-and-job-owned namespace, so separate Agents, jobs, and reverse directions can coexist without reconciliation deleting one another’s active objects.
+
 ## Supported releases
 
 Netdata's built-in Ceph profile recognizes the metric surfaces of:
@@ -184,7 +225,9 @@ The built-in Ceph profile recognizes all three Prometheus interfaces. Alert owne
 - placement group, pool, capacity, and recovery conditions;
 - node-proxy hardware conditions exposed by MGR;
 - gateway-local NVMe-oF conditions exposed by each gateway-exporter job;
-- RGW notification, Lua, request-fallback, queue-pressure, and multisite retry conditions.
+- RGW notification, Lua, request-fallback, queue-pressure, and multisite retry conditions;
+- authenticated S3 stage and multisite phase failures, plus configured latency objectives from each `s3check` job;
+- directional multisite payload mismatches, RPO breaches, and delete-propagation objectives.
 
 The native Dashboard collector owns API component collection failures. Generic Netdata collectors own host-local and endpoint checks:
 
@@ -215,7 +258,10 @@ The MGR profile provides aggregate RGW telemetry for requests, aborted requests,
 
 Collect the RGW JSON access log with `web_log` to analyze HTTP outcomes, bytes, clients, and latency. Configure RGW to emit its access log in JSON format and make that file available to the Agent. The Ceph JSON example maps request, status, size, and client fields, and declares `total_time` as a numeric custom field in milliseconds, preserving Ceph's exact duration field.
 
-Use `httpcheck` for unauthenticated endpoint liveness and `x509check` for certificate expiration or revocation.
+Use `s3check` for authenticated object lifecycle correctness, client-vantage latency, and directional multisite
+replication. The MGR multisite counters show RGW replication work and retries; `s3check` proves what a client can
+currently read at the destination and whether the payload is identical. Keep `httpcheck` for unauthenticated endpoint
+liveness and `x509check` for certificate expiration or revocation.
 
 ## Investigation Functions
 
@@ -272,6 +318,7 @@ For collector configuration details, see:
 
 - [Ceph](/docs/collecting-metrics/collectors/storage-and-filesystems/ceph)
 - [Ceph Prometheus](/docs/collecting-metrics/collectors/storage-and-filesystems/ceph-prometheus)
+- [S3 Compatible Object Storage](/docs/collecting-metrics/collectors/synthetic-testing/s3-compatible-object-storage)
 
 For Agent deployment, streaming, retention, exporting, and notifications, see:
 

diff --git a/docs/Collecting Metrics/Collectors/Collectors Page 2.mdx b/docs/Collecting Metrics/Collectors/Collectors Page 2.mdx
@@ -281,6 +281,9 @@ import { Grid, Box, GridPagination } from '@site/src/components/Grid_integration
 <Box banner="by Netdata" banner_color="#00ab44" to="/docs/collecting-metrics/collectors/storage-and-filesystems/s.m.a.r.t." title="S.M.A.R.T.">
     <img custom-image src="https://netdata.cloud/img/smart.png" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="ok" data-logo-contrast-dark="ok" data-logo-contrast-confidence="high"/>
 </Box>
+<Box banner="by Netdata" banner_color="#00ab44" to="/docs/collecting-metrics/collectors/synthetic-testing/s3-compatible-object-storage" title="S3 Compatible Object Storage">
+    <img custom-image src="https://netdata.cloud/img/synthetics.svg" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="unknown" data-logo-contrast-dark="unknown" data-logo-contrast-confidence="low"/>
+</Box>
 <Box banner="by Netdata" banner_color="#00ab44" to="/docs/collecting-metrics/collectors/storage-and-filesystems/samba" title="Samba">
     <img custom-image src="https://netdata.cloud/img/samba.svg" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="low" data-logo-contrast-dark="ok" data-logo-contrast-confidence="medium"/>
 </Box>
@@ -539,9 +542,6 @@ import { Grid, Box, GridPagination } from '@site/src/components/Grid_integration
 <Box banner="by Community" banner_color="rgba(0, 0, 0, 0.25)" to="/docs/collecting-metrics/collectors/containers-and-vms/cilium-operator" title="Cilium Operator">
     <img custom-image src="https://netdata.cloud/img/cilium.png" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="ok" data-logo-contrast-dark="ok" data-logo-contrast-confidence="high"/>
 </Box>
-<Box banner="by Community" banner_color="rgba(0, 0, 0, 0.25)" to="/docs/collecting-metrics/collectors/containers-and-vms/cilium-proxy" title="Cilium Proxy">
-    <img custom-image src="https://netdata.cloud/img/cilium.png" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="ok" data-logo-contrast-dark="ok" data-logo-contrast-confidence="high"/>
-</Box>
 </Grid>
 
 <GridPagination basePath="/docs/collecting-metrics/collectors" currentPage={2} pageCount={3} />
diff --git a/docs/Collecting Metrics/Collectors/Collectors Page 3.mdx b/docs/Collecting Metrics/Collectors/Collectors Page 3.mdx
@@ -17,6 +17,9 @@ description: "Browse Netdata integrations for Collecting Metrics / Collectors an
 import { Grid, Box, GridPagination } from '@site/src/components/Grid_integrations';
 
 <Grid columns="4">
+<Box banner="by Community" banner_color="rgba(0, 0, 0, 0.25)" to="/docs/collecting-metrics/collectors/containers-and-vms/cilium-proxy" title="Cilium Proxy">
+    <img custom-image src="https://netdata.cloud/img/cilium.png" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="ok" data-logo-contrast-dark="ok" data-logo-contrast-confidence="high"/>
+</Box>
 <Box banner="by Community" banner_color="rgba(0, 0, 0, 0.25)" to="/docs/collecting-metrics/collectors/applications/clamav-daemon" title="ClamAV daemon">
     <img custom-image src="https://netdata.cloud/img/clamav.png" style={{width: '90%', maxHeight: '100%', verticalAlign: 'middle' }} alt="" data-integration-logo="true" data-logo-contrast-light="ok" data-logo-contrast-dark="ok" data-logo-contrast-confidence="high"/>
 </Box>

diff --git a/...ctors/Storage and Filesystems/Physical and Logical Disk Performance Metrics.mdx b/...ctors/Storage and Filesystems/Physical and Logical Disk Performance Metrics.mdx
@@ -73,6 +73,7 @@ No action required.
 | Option | Description | Default | Required |
 |:-----|:------------|:--------|:---------:|
 | update every | Data collection frequency. | 1 | no |
+| exclude space metrics on paths | Space-separated, case-insensitive Netdata simple patterns for logical-volume mount points to exclude from disk space collection. Patterns support `*` wildcards and `!` negative terms. For example, `*AssuredRecoveryTemp*` excludes temporary backup volumes with that path component. |  | no |
 
 
 
@@ -101,7 +102,20 @@ sudo ./edit-config netdata.conf
 ```
 
 ##### Examples
-There are no configuration examples.
+
+###### Exclude temporary backup volumes
+
+Exclude logical-volume disk space metrics for mount points that contain `AssuredRecoveryTemp`.
+
+<details open>
+<summary>Configuration</summary>
+
+```yaml
+[plugin:windows:PerflibStorage]
+  exclude space metrics on paths = *AssuredRecoveryTemp*
+
+```
+</details>
 
 
 
@@ -142,9 +156,10 @@ Labels:
 | Label      | Description     |
 |:-----------|:----------------|
 | mount_point | Drive letter or mount point path assigned by Windows (e.g., 'C:', 'D:'). |
-| driver_type | Classification of the disk device (e.g., norootdir, removable, cdrom, ramdisk). |
+| drive_type | Classification of the disk device (e.g., norootdir, removable, cdrom, ramdisk). |
 | filesystem | File system format used on the volume (e.g., NTFS, FAT32). |
 | rw_mode | Current read/write permissions status of the volume (read-only access, read and write access). |
+| serial_number | Volume serial number in hexadecimal format. |
 
 Metrics: