- [Use the v3 lightweight client libraries](#use-the-v3-lightweight-client-libraries)
- [Avoid features that are not included with InfluxDB Cloud Dedicated](#avoid-features-that-are-not-included-with-influxdb-cloud-dedicated)
- [Use SQL or InfluxQL as your Query Language](#use-sql-or-influxql-as-your-query-language)
- [Stay within the schema limits of InfluxDB Cloud Serverless](#stay-within-the-schema-limits-of-influxdb-cloud-serverless)
- [Keep test and production data separate](#keep-test-and-production-data-separate)
<!-- END TOC -->
## Key differences between InfluxDB Cloud Serverless and Cloud Dedicated
### User interface differences
The user interfaces between InfluxDB Cloud Serverless and InfluxDB Cloud
Dedicated are completely different.
InfluxDB Cloud Serverless is designed for individual developers, data scientists,
and general hobbyists, as well as business-to-business (B2B) customers.
The Cloud Serverless graphical user interface (GUI) provides basic features for database administration (add/delete databases, generate tokens, etc.), query, visualization, and dashboarding.
Unlike Cloud Serverless, Cloud Dedicated does not come with a GUI.
Cloud Dedicated customers use an administrative command line tool (`influxctl`)
for managing databases and tokens. The `influxctl` utility is not available for
InfluxDB Cloud Serverless. Because the platforms use different administrative tools,
if you're using Cloud Serverless as an evaluation platform for Cloud Dedicated,
you won’t be able to evaluate the Cloud Dedicated administrative features directly.
### Terminology differences
InfluxDB Cloud Serverless was an upgrade that introduced the InfluxDB 3.0 storage
engine to InfluxData’s original InfluxDB Cloud (TSM) multi-tenant solution.
InfluxDB Cloud utilizes the Time-Structured Merge Tree (TSM) storage engine in
which databases were referred to as "buckets".
Cloud Serverless still uses this term.
InfluxDB Cloud Dedicated has only ever used the InfluxDB 3.0 storage engine.
Resource names in Cloud Dedicated are more traditionally aligned with SQL database engines.
Databases are named "databases" and "measurements" are structured as tables.
"Tables" or "measurements" can be used interchangeably.
The v2 API (which uses the Flux language for querying) is reachable in InfluxDB
Cloud Serverless, but isn't supported. If you plan to use InfluxDB Cloud
Serverless as an evaluation or staging platform for InfluxDB Cloud Dedicated,
use SQL or InfluxQL.
### API and client library differences
Because the v2 Query API uses Flux, customers should avoid using the v2 API when
querying InfluxDB Cloud Serverless as an evaluation, staging, or prototyping
platform for InfluxDB Cloud Dedicated.
For writing data, InfluxDB Cloud Dedicated and InfluxDB Cloud Serverless both
support the v1 API and the v2 write API.
In addition, [InfluxDB v3 client libraries](/influxdb/reference/client-libraries/v3/)
are available that work the same for both InfluxDB Cloud Serverless and InfluxDB
Cloud Dedicated and help avoid any API differences between the two platforms.
For more detailed information about choosing a client library, see the
_[Choosing a Client Library When Developing with InfluxDB 3.0](https://www.influxdata.com/blog/choosing-client-library-when-developing-with-influxdb-3-0/)_
blog post.
### Tasks and alerts differences
InfluxDB Cloud Serverless had built-in task and alert systems.
These were built on Flux and have not been carried forward to InfluxDB Cloud Dedicated.
If you use InfluxDB Cloud Serverless as an evaluation platform for
InfluxDB Cloud Dedicated, don’t utilize these features as they aren't available
on InfluxDB Cloud Dedicated.
With InfluxDB Cloud Dedicated, you can build custom task and alerting solutions or use 3rd-party tools like Grafana or Prefect--for example:
- [Send alerts using data in InfluxDB Cloud Serverless](/influxdb/cloud-serverless/process-data/send-alerts/)
| API token generation | Built-in UI or v2 API | `influxctl` CLI |
### Performance differences
InfluxDB Cloud Serverless is a multi-tenant solution and has a number of factors
that could affect database performance, including:
- **Rate limiting**: Limits on reads and writes.
- **Noisy neighbors**: Unexpected high usage from other customers could impact your experience.
- **Generic optimization**: Cloud Serverless is generically optimized to fit the
most common workloads. InfluxDB Cloud Dedicated can be tuned to your
specific data sets and workload with options such as custom partitioning.
- **Tag-specific queries**: Due to generic optimization, queries that select rows with a subset
of tag values may be noticeably slower than in previous versions.
If you are using InfluxDB Cloud Serverless for evaluation purposes, consider
investing in an InfluxDB Cloud Dedicated proof-of-concept (PoC) to get
a more accurate picture of your expected performance.
### Schema differences
Schema support in InfluxDB Cloud Serverless and InfluxDB Cloud Dedicated is nearly identical.
However, InfluxDB Cloud Dedicated provides more flexibility, letting you raise
or exceed limits on the number of tables (measurements) and columns
(time, fields, and tags) in a database.
_See [Stay within the schema limits of InfluxDB Cloud Serverless](#stay-within-the-schema-limits-of-influxdb-cloud-serverless)_.
### Organization management differences
InfluxDB Cloud Serverless has a feature that lets you create sub-organizations
within your account. InfluxDB Cloud Dedicated does not have this feature.
Additionally, due to being multi-tenant, Cloud Serverless requires specifying
organization name or ID in some interactions. This isn't required with Cloud Dedicated.
## Best practices
Follow these recommended best practices when using InfluxDB Cloud Serverless
as an evaluation or prototyping platform for InfluxDB Cloud Dedicated.
### Use the v3 lightweight client libraries
Use the InfluxDB [v3 lightweight client libraries](/influxdb/cloud-serverless/reference/client-libraries/v3/)
to help make your code for writing and querying cross-compatible with InfluxDB Cloud Serverless and Cloud Dedicated.
You'll only need to change your InfluxDB connection credentials
(host, database name, and token).
### Avoid features that are not included with InfluxDB Cloud Dedicated
The easiest way to avoid using features in InfluxDB Cloud Serverless that don’t
exist in Cloud Dedicated is to avoid using the Cloud Serverless UI, except when
managing tokens and buckets.
In order to maintain compatibility with Cloud Dedicated, specifically avoid using the following
InfluxDB Cloud Serverless features:
- The v2 query API and the Flux language
- Administrative APIs
- Tasks and alerts from the Cloud Serverless UI (instead use one of the options
mentioned in _[Tasks and alerts differences](#tasks-and-alerts-differences)_).
- InfluxDB dashboards and visualization tools (use 3rd-party visualization tools)
### Use SQL or InfluxQL as your Query Language
SQL and InfluxQL are optimized for InfluxDB v3 and both are excellent
options in Cloud Dedicated and Cloud Serverless.
Avoid Flux since it can’t be used with InfluxDB Cloud Dedicated.
### Stay within the schema limits of InfluxDB Cloud Serverless
If you stay within InfluxDB Cloud Serverless limits for tables (measurements)
and columns (time, fields, and tags) within a table, then you won’t have any
problems with limits in InfluxDB Cloud Dedicated.
Cloud Dedicated also provides more flexibility by letting you configure limits.
| Description | Limit |
| :--------------------------- | ----: |
| Tables (measurements) | 500 |
| Columns (time, fields, tags) | 200 |
### Keep test and production data separate
If you are using InfluxDB Cloud Serverless for prototyping or evaluation, use
test data and don’t store production data on the platform.
This has the following benefits:
- Using test data is less risky. If you make a mistake with your schema, you can delete your data and start over. If you use production
data that you rely on and only have one copy, schema mistakes are harder to
recover from.
- You won't need to migrate data. After you have completed your evaluation, you can delete your test data and start using Cloud Dedicated.
_You can use live or production data for protoype in InfluxDB Cloud Serverless, and then migrate it to Cloud Dedicated after you complete your evaluation. However, currently this requires
that you write code to query and extract the data, convert it to line protocol
format, and then write it to InfluxDB Cloud Dedicated._
Scott Anderson
GitHub