Scale InfluxDB Clustered (#5526)

* clustered scaling guide

* Update content/influxdb/clustered/admin/scale-cluster.md

Co-authored-by: Jason Stirnaman <jstirnaman@influxdata.com>

* add information about ingester storage volumes to scaling guide

* Update content/influxdb/clustered/reference/glossary.md

Co-authored-by: Jason Stirnaman <jstirnaman@influxdata.com>

---------

Co-authored-by: Jason Stirnaman <jstirnaman@influxdata.com>
pull/5527/head
Scott Anderson 2024-07-22 10:05:53 -06:00 committed by GitHub
parent 61b127870f
commit e851b0089e
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
5 changed files with 385 additions and 70 deletions

View File

@ -7,7 +7,7 @@ description: >
menu:
influxdb_clustered:
parent: Administer InfluxDB Clustered
weight: 207
weight: 208
---
{{< product-name >}} generates a valid access token (known as the _admin token_)

View File

@ -0,0 +1,352 @@
---
title: Scale your InfluxDB cluster
description: >
InfluxDB Clustered lets you scale individual components of your cluster both
vertically and horizontally to match your specific workload.
menu:
influxdb_clustered:
parent: Administer InfluxDB Clustered
name: Scale your cluster
weight: 207
influxdb/clustered/tags: [scale]
related:
- /influxdb/clustered/reference/internals/storage-engine/
- https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#requests-and-limits, Kubernetes resource requests and limits
---
InfluxDB Clustered lets you scale individual components of your cluster both
vertically and horizontally to match your specific workload.
Use the `AppInstance` resource defined in your `influxdb.yml` to manage
resources available to each component.
- [Scaling strategies](#scaling-strategies)
- [Vertical scaling](#vertical-scaling)
- [Horizontal scaling](#horizontal-scaling)
- [Scale components in your cluster](#scale-components-in-your-cluster)
- [Horizontally scale a component](#horizontally-scale-a-component)
- [Vertically scale a component](#vertically-scale-a-component)
- [Apply your changes](#apply-your-changes)
- [Scale your cluster as a whole](#scale-your-cluster-as-a-whole)
- [Recommended scaling strategies per component](#recommended-scaling-strategies-per-component)
- [Ingester](#ingester)
- [Querier](#querier)
- [Router](#router)
- [Compactor](#compactor)
- [Catalog](#catalog)
- [Object store](#object-store)
## Scaling strategies
The following scaling strategies can be applied to components in your InfluxDB
cluster.
### Vertical scaling
Vertical scaling (also known as "scaling up") involves increasing the resources
(such as RAM or CPU) available to a process or system.
Vertical scaling is typically used to handle resource-intensive tasks that
require more processing power.
{{< html-diagram/scaling-strategy "vertical" >}}
### Horizontal scaling
Horizontal scaling (also known as "scaling out") involves increasing the number of
nodes or processes available to perform a given task.
Horizontal scaling is typically used to increase the amount of workload or
throughput a system can manage, but also provides additional redundancy and failover.
{{< html-diagram/scaling-strategy "horizontal" >}}
## Scale components in your cluster
The following components of your InfluxDB cluster are scaled by modifying
properties in your `AppInstance` resource:
- Ingester
- Querier
- Compactor
- Router
{{% note %}}
#### Scale your Catalog and Object store
Your InfluxDB [Catalog](/influxdb/clustered/reference/internals/storage-engine/#catalog)
and [Object store](/influxdb/clustered/reference/internals/storage-engine/#object-store)
are managed outside of your `AppInstance` resource. Scaling mechanisms for these
components depend on the technology and underlying provider used for each.
{{% /note %}}
Use the `spec.package.spec.resources` property in your `AppInstance` resource
defined in your `influxdb.yml` to define system resource minimums and limits
for each pod and the number of replicas per component.
`requests` are the minimum that the Kubernetes scheduler should reserve for a pod.
`limits` are the maximum that a pod should be allowed to use.
Your `AppInstance` resource can include the following properties to define
resource minimums and limits per pod and replicas per component:
- `spec.package.spec.resources`
- `ingester`
- `requests`
- `cpu`: Minimum CPU resource units to assign to ingesters
- `memory`: Minimum memory resource units to assign to ingesters
- `replicas`: Number of ingester replicas to provision
- `limits`
- `cpu`: Maximum CPU resource units to assign to ingesters
- `memory`: Maximum memory resource units to assign to ingesters
- `compactor`
- `requests`
- `cpu`: Minimum CPU resource units to assign to compactors
- `memory`: Minimum memory resource units to assign to compactors
- `replicas`: Number of compactor replicas to provision
- `limits`
- `cpu`: Maximum CPU resource units to assign to compactors
- `memory`: Maximum memory resource units to assign to compactors
- `querier`
- `requests`
- `cpu`: Minimum CPU resource units to assign to queriers
- `memory`: Minimum memory resource units to assign to queriers
- `replicas`: Number of querier replicas to provision
- `limits`
- `cpu`: Maximum CPU resource units to assign to queriers
- `memory`: Maximum memory resource units to assign to queriers
- `router`
- `requests`
- `cpu`: Minimum CPU resource units to assign to routers
- `memory`: Minimum memory resource units to assign to routers
- `replicas`: Number of router replicas to provision
- `limits`
- `cpu`: Maximum CPU Resource units to assign to routers
- `memory`: Maximum memory resource units to assign to routers
{{< expand-wrapper >}}
{{% expand "View example `AppInstance` with resource requests and limits" %}}
{{% code-placeholders "(INGESTER|COMPACTOR|QUERIER|ROUTER)_(CPU_(MAX|MIN)|MEMORY_(MAX|MIN)|REPLICAS)" %}}
```yml
apiVersion: kubecfg.dev/v1alpha1
kind: AppInstance
# ...
spec:
package:
spec:
# The following settings tune the various pods for their cpu/memory/replicas
# based on workload needs. Only uncomment the specific resources you want
# to change. Anything left commented will use the package default.
resources:
ingester:
requests:
cpu: INGESTER_CPU_MIN
memory: INGESTER_MEMORY_MIN
replicas: INGESTER_REPLICAS # Default is 3
limits:
cpu: INGESTER_CPU_MAX
memory: INGESTER_MEMORY_MAX
compactor:
requests:
cpu: COMPACTOR_CPU_MIN
memory: COMPACTOR_MEMORY_MIN
replicas: COMPACTOR_REPLICAS # Default is 1
limits:
cpu: COMPACTOR_CPU_MAX
memory: COMPACTOR_MEMORY_MAX
querier:
requests:
cpu: QUERIER_CPU_MIN
memory: QUERIER_MEMORY_MIN
replicas: QUERIER_REPLICAS # Default is 1
limits:
cpu: QUERIER_CPU_MAX
memory: QUERIER_MEMORY_MAX
router:
requests:
cpu: ROUTER_CPU_MIN
memory: ROUTER_MEMORY_MIN
replicas: ROUTER_REPLICAS # Default is 1
limits:
cpu: ROUTER_CPU_MAX
memory: ROUTER_MEMORY_MAX
```
{{% /code-placeholders %}}
{{% /expand %}}
{{< /expand-wrapper >}}
{{% note %}}
Applying resource limits to pods is optional, but provides better resource
isolation and protects against pods using more resources than intended. For
information, see [Kubernetes resource requests and limits](https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#requests-and-limits).
{{% /note %}}
##### Related Kubernetes documentation
- [CPU resource units](https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#meaning-of-cpu)
- [Memory resource units](https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#meaning-of-memory)
### Horizontally scale a component
To horizontally scale a component in your InfluxDB cluster, increase or decrease
the number of replicas for the component in the `spec.package.spec.resources`
property in your `AppInstance` resource and [apply the change](#apply-your-changes).
{{% warn %}}
#### Only use the AppInstance to scale component replicas
Only use the `AppInstance` resource to scale component replicas.
Manually scaling replicas may cause errors.
{{% /warn %}}
For example--to horizontally scale your
[Ingester](/influxdb/clustered/reference/internals/storage-engine/#ingester):
```yaml
apiVersion: kubecfg.dev/v1alpha1
kind: AppInstance
# ...
spec:
package:
spec:
resources:
ingester:
requests:
# ...
replicas: 6
```
### Vertically scale a component
To vertically scale a component in your InfluxDB cluster, increase or decrease
the CPU and memory resource units to assign to component pods in the
`spec.package.spec.resources` property in your `AppInstance` resource and
[apply the change](#apply-your-changes).
```yaml
apiVersion: kubecfg.dev/v1alpha1
kind: AppInstance
# ...
spec:
package:
spec:
resources:
ingester:
requests:
cpu: "500m"
memory: "512MiB"
# ...
limits:
cpu: "1000m"
memory: "1024MiB"
```
### Apply your changes
After modifying the `AppInstance` resource, use `kubectl apply` to apply the
configuration changes to your cluster and scale the updated components.
```sh
kubectl apply \
--filename myinfluxdb.yml \
--namespace influxdb
```
## Scale your cluster as a whole
Scaling your entire InfluxDB Cluster is done by scaling your Kubernetes cluster
and is managed outside of InfluxDB. The process of scaling your entire Kubernetes
cluster depends on your underlying Kubernetes provider. You can also use
[Kubernetes autoscaling](https://kubernetes.io/docs/concepts/cluster-administration/cluster-autoscaling/)
to automatically scale your cluster as needed.
## Recommended scaling strategies per component
- [Ingester](#ingester)
- [Querier](#querier)
- [Router](#router)
- [Compactor](#compactor)
- [Catalog](#catalog)
- [Object store](#object-store)
### Ingester
The Ingester can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
Vertical scaling increases write throughput and is typically the most effective
scaling strategy for the Ingester.
#### Ingester storage volume
Ingesters use an attached storage volume to store the
[Write-Ahead Log (WAL)](/influxdb/clustered/reference/glossary/#wal-write-ahead-log).
With more storage available, Ingesters can keep bigger WAL buffers, which
improves query performance and reduces pressure on the Compactor.
Storage speed also helps with query performance.
Configure the storage volume attached to Ingester pods in the
`spec.package.spec.ingesterStorage` property of your `AppInstance` resource.
{{< expand-wrapper >}}
{{% expand "View example Ingester storage configuration" %}}
{{% code-placeholders "STORAGE_(CLASS|SIZE)" %}}
```yml
apiVersion: kubecfg.dev/v1alpha1
kind: AppInstance
# ...
spec:
package:
spec:
# ...
ingesterStorage:
# (Optional) Set the storage class. This will differ based on the K8s
#environment and desired storage characteristics.
# If not set, the default storage class is used.
storageClassName: STORAGE_CLASS
# Set the storage size (minimum 2Gi recommended)
storage: STORAGE_SIZE
```
{{% /code-placeholders %}}
{{% /expand %}}
{{< /expand-wrapper >}}
### Querier
The Querier can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
Horizontal scaling increases query throughput to handle more concurrent queries.
Vertical scaling improves the Queriers ability to process computationally
intensive queries.
### Router
The Router can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
Horizontal scaling increases request throughput and is typically the most effective
scaling strategy for the Router.
### Compactor
The Compactor can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
Because compaction is a compute-heavy process, vertical scaling (especially
increasing the available CPU) is the most effective scaling strategy for the
Compactor. Horizontal scaling increases compaction throughput, but not as
efficiently as vertical scaling.
### Catalog
Scaling strategies available for the Catalog depend on the PostgreSQL-compatible
database used to run the catalog. All support [vertical scaling](#vertical-scaling).
Most support [horizontal scaling](#horizontal-scaling) for redundancy and failover.
### Object store
Scaling strategies available for the Object store depend on the underlying
object storage services used to run the object store. Most support
[horizontal scaling](#horizontal-scaling) for redundancy, failover, and
increased capacity.

View File

@ -757,21 +757,21 @@ your anticipated workload.
in your `myinfluxdb.yml`. If omitted, your cluster will use the default scale settings.
- `spec.package.spec.resources`
- `.ingester.requests`
- `.cpu`: CPU Resource units to assign to ingesters
- `.memory`: Memory resource units to assign to ingesters
- `ingester.requests`
- `cpu`: CPU resource units to assign to ingesters
- `memory`: Memory resource units to assign to ingesters
- `replicas`: Number of ingester replicas to provision
- `.compactor.requests`
- `.cpu`: CPU Resource units to assign to compactors
- `.memory`: Memory resource units to assign to compactors
- `compactor.requests`
- `cpu`: CPU resource units to assign to compactors
- `memory`: Memory resource units to assign to compactors
- `replicas`: Number of compactor replicas to provision
- `.querier.requests`
- `.cpu`: CPU Resource units to assign to queriers
- `.memory`: Memory resource units to assign to queriers
- `querier.requests`
- `cpu`: CPU resource units to assign to queriers
- `memory`: Memory resource units to assign to queriers
- `replicas`: Number of querier replicas to provision
- `.router.requests`
- `.cpu`: CPU Resource units to assign to routers
- `.memory`: Memory resource units to assign to routers
- `router.requests`
- `cpu`: CPU resource units to assign to routers
- `memory`: Memory resource units to assign to routers
- `replicas`: Number of router replicas to provision
###### Related Kubernetes documentation

View File

@ -1131,15 +1131,17 @@ A statement that sets or updates the value stored in a variable.
## W
### WAL (Write Ahead Log) - enterprise
### WAL (Write-Ahead Log)
The temporary cache for recently written points.
To reduce the frequency that permanent storage files are accessed, InfluxDB
caches new points in the WAL until their total size or age triggers a flush to
more permanent storage. This allows for efficient batching of the writes into the TSM.
more permanent storage. This allows for efficient batching of the writes into
the storage engine.
Points in the WAL can be queried and persist through a system reboot.
On process start, all points in the WAL must be flushed before the system accepts new writes.
Points in the WAL are queryable and persist through a system reboot.
On process start, all points in the WAL must be flushed before the system
accepts new writes.
Related entries:
[tsm](#tsm-time-structured-merge-tree)

View File

@ -28,9 +28,6 @@ queries, and is optimized to reduce storage cost.
- [Catalog](#catalog)
- [Object store](#object-store)
- [Compactor](#compactor)
- [Scaling strategies](#scaling-strategies)
- [Vertical scaling](#vertical-scaling)
- [Horizontal scaling](#horizontal-scaling)
## Storage engine diagram
@ -66,8 +63,8 @@ In this process, the Ingester does the following:
##### Ingester scaling strategies
The Ingester can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
The Ingester can be scaled both [vertically](/influxdb/clustered/admin/scale-cluster/#vertical-scaling)
and [horizontally](/influxdb/clustered/admin/scale-cluster/#horizontal-scaling).
Vertical scaling increases write throughput and is typically the most
effective scaling strategy for the Ingester.
@ -98,10 +95,11 @@ At query time, the querier:
##### Querier scaling strategies
The Querier can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
The Querier can be scaled both [vertically](/influxdb/clustered/admin/scale-cluster/#vertical-scaling)
and [horizontally](/influxdb/clustered/admin/scale-cluster/#horizontal-scaling).
Horizontal scaling increases query throughput to handle more concurrent queries.
Vertical scaling improves the Querier's ability to process computationally intensive queries.
Vertical scaling improves the Querier's ability to process computationally
intensive queries.
### Catalog
@ -117,8 +115,10 @@ It fulfills the following roles:
##### Catalog scaling strategies
Scaling strategies available for the Catalog depend on the PostgreSQL-compatible
database used to run the catalog. All support [vertical scaling](#vertical-scaling).
Most support [horizontal scaling](#horizontal-scaling) for redundancy and failover.
database used to run the catalog. All support
[vertical scaling](/influxdb/clustered/admin/scale-cluster/#vertical-scaling).
Most support [horizontal scaling](/influxdb/clustered/admin/scale-cluster/#horizontal-scaling)
for redundancy and failover.
### Object store
@ -132,8 +132,8 @@ Data in each Parquet file is sorted, encoded, and compressed.
Scaling strategies available for the Object store depend on the underlying
object storage services used to run the object store.
Most support [horizontal scaling](#horizontal-scaling) for redundancy, failover,
and increased capacity.
Most support [horizontal scaling](/influxdb/clustered/admin/scale-cluster/#horizontal-scaling)
for redundancy, failover, and increased capacity.
### Compactor
@ -143,48 +143,9 @@ It then updates the [Catalog](#catalog) with locations of compacted data.
##### Compactor scaling strategies
The Compactor can be scaled both [vertically](#vertical-scaling) and
[horizontally](#horizontal-scaling).
The Compactor can be scaled both [vertically](/influxdb/clustered/admin/scale-cluster/#vertical-scaling)
and [horizontally](/influxdb/clustered/admin/scale-cluster/#horizontal-scaling).
Because compaction is a compute-heavy process, vertical scaling (especially
increasing the available CPU) is the most effective scaling strategy for the Compactor.
Horizontal scaling increases compaction throughput, but not as efficiently as
vertical scaling.
---
## Scaling strategies
The following scaling strategies can be applied to components of the InfluxDB v3
storage architecture.
{{% note %}}
<!-- Cloud Dedicated-specific -->
For information about scaling your {{< product-name >}} infrastructure,
[contact InfluxData support](https://support.influxdata.com).
{{% /note %}}
### Vertical scaling
Vertical scaling (also known as "scaling up") involves increasing the resources
(such as RAM or CPU) available to a process or system.
Vertical scaling is typically used to handle resource-intensive tasks that
require more processing power.
{{< html-diagram/scaling-strategy "vertical" >}}
### Horizontal scaling
Horizontal scaling (also known as "scaling out") involves increasing the number of
nodes or processes available to perform a given task.
Horizontal scaling is typically used to increase the amount of workload or
throughput a system can manage, but also provides additional redundancy and failover.
{{% warn %}}
#### Only use the AppInstance to scale component replicas
Manually scaling resources may cause errors.
{{% /warn %}}
{{< html-diagram/scaling-strategy "horizontal" >}}