Manage data storage and throughput limits

To ensure good performance, the HERE platform has limits on data storage and throughput. Some limits can be controlled by layer configuration, which may impact your cost since you are charged based on how you have configured the layers and data usage. As a general rule, the more data you send and receive from the HERE platform, and the more data you store, the more you will be charged. You will also be charged more if you configure layers for higher performance.

For more information on costs associated with each layer type, see the Cost Management Guide.

For more information on pipeline logging, see Pipeline logging.

Catalog limits

Constraint	Limit
Maximum layers in a catalog	250
Automatic catalog version deletion (maximum number of retainable versions)	50,000

Layer limits

Storage and throughput limits

Index layer limits

Constraint	Limit
Maximum partition size	500 GB
Maximum throughput	No established limit, but your throughput will be affected by these factors: Load on the HERE platform globally Load on your HERE platform org Type of operation (for example, read, write, or delete) Size of the request or batch Number of concurrent requests
Minimum TTL	7 days
Maximum TTL	No maximum
Maximum amount of data	No maximum
Minimum number of indexes of an index layer	1
Maximum number of indexes of an index layer	4

Cost considerations:

You are charged for the amount of metadata and data stored in the layer
You are charged based on the amount of data you read and write using the blob and index APIs
You are charged based on the retention setting configured for the layer

Interactive map layer limits

Constraint	Limit
Maximum partition size	Not applicable
Maximum throughput	No established limit, but your throughput will be affected by these factors: Load on the HERE platform globally Load on your HERE platform org Type of operation (for example, read, write, or delete) Size of the request or batch Number of concurrent requests
Minimum TTL	Not applicable
Maximum TTL	Not applicable
Maximum amount of data	128 TB
Maximum payload size - uploads	20 MB (uncompressed)
Maximum payload size - downloads	50 MB (uncompressed)
Maximum requests	If you exceed 1,000 requests in a second, you are throttled and the remaining requests within that second fail for each app.
Maximum duration of a single request	30 seconds
Maximum features in a response	100,000
Maximum features in a layer to be fully searchable	There are no restrictions for property search in layers that contain less than 10,000 features. For layers that contain more than 10,000 features, property search is enabled only for properties marked as “searchable”. Searchable properties can be defined either automatically or manually using the layer configuration settings.

Cost considerations:

You are charged for the amount of metadata and data stored in the layer
You are charged based on the amount of data you read and write using the interactive API
Indexing of layers happens on the fly, so no additional IO is charged, but the indexes add to the data stored.

Object store layer limits

Constraint	Limit
Maximum single object size	500 GB
Maximum throughput	No established limit, but your throughput will be affected by these factors: Load on the HERE platform globally Load on your HERE platform org Type of operation (for example, read, write, or delete) Size of the request or batch Number of concurrent requests
Minimum TTL	Not applicable
Maximum TTL	Not applicable
Maximum amount of data in an object store layer	No limit

Cost considerations:

You are charged for the amount of metadata and data stored in the layer
You are charged based on the amount of data you read and write using the blob API
There are no layer configuration settings that affect cost

Stream layer limits

Constraint	Limit
Maximum partition size	1 MB
Maximum throughput	Inbound: 32800 KBps Outbound: 65500 KBps, but catalogs in the HERE Marketplace have a maximum outbound throughput of 2000 KBps. The maximum throughput for a stream layer is configured as allocated capacity during layer creation. The service starts throttling inbound messages when the inbound rate exceeds the inbound throughput. The service starts throttling outbound messages when the total outbound rate to all consumers exceeds the outbound throughput. When throttling occurs, the service response is delayed, but no messages are dropped.
Minimum TTL	10 minutes
Maximum TTL	72 hours
Maximum amount of data	The maximum amount of data that can be concurrently stored in a stream layer can be calculated as follows: `(inbound throughput) x (retention time)`
Maximum consumer group size	The number of subscriptions in a consumer group in a stream layer can be calculated as follows: `max (input throughput, ceil (output throughput /2))`
Maximum message size	The maximum message size for a stream layer is 1 MB, and we recommend that you keep messages to 1 MB or less. For messages larger than 1 MB, we recommend that you upload the data to the Blob API first and pass a message in stream by reference (data handle).

Cost considerations:

Usage does not affect cost
You are charged based on the maximum throughput configured for the layer
You are charged based on the retention setting configured for the layer

Versioned layer limits

Constraint	Limit
Maximum partition size	500 GB
Maximum throughput	We recommend that you don't exceed 24 MB/second when publishing metadata. More specific limits apply for single file uploads in the data publication process. For more information see Publish to a versioned layer.
Maximum amount of data	No maximum

Cost considerations:

HERE charges you for the amount of metadata and data stored in the layer
HERE charges you based on the amount of data you read and write using the blob and metadata APIs
There are no layer configuration settings that affect cost
Deleting catalog versions to manage storage costs gives you more granular data lifecycle management controls for your versioned layers. You can delete catalog versions manually or use the config service to delete catalog versions automatically by enabling the automaticVersionDeletion and setting the numberOfVersionsToKeep with a maximum value of 50,000. For more information, see Delete catalog versions.
HERE recommends your data partitions have homogenous sizes and use as high of a zoom level as possible. For more information, see Considerations when designing data models for versioned layers.

Volatile layer limits

Constraint	Limit
Maximum partition size	2 MB
Maximum throughput	No established limit, but your throughput will be affected by these factors: Load on the HERE platform globally Load on your HERE platform org Type of operation (for example, read, write, or delete) Size of the request or batch Number of concurrent requests
Minimum TTL	1 minute
Maximum TTL	7 days
Maximum amount of data	21 GB

Cost considerations:

You are charged based on the amount of data you read and write using the blob API
You are charged based on the storage capacity configured for the layer. The price increases as the storage capacity increases.
You are charged based on the data redundancy configured for the layer. For multi-instance mode, your cost triples because two additional copies of your data are created, for a total of 3 copies, to provide durability.

Data API quotas

API	Quota
`stream`	Each `appId` has a quota for total stream layer throughput. Quotas are byte-rate thresholds. A single `appId` can span multiple producer and consumer instances, and the quota will apply for all of them as a single entity. For example, if the `appId` ”test-client” has a produce quota of 10200 KBps, this is shared across all instances with that same `appId`. The default quota per Kafka topic partition is: Input throughput ("producer_byte_rate"): 10200 KBps Output throughput rate ("consumer_byte_rate"): 2000 KBps
`index`	The `index` API uses a resource-based throttling mechanism. When requests cause CPU and memory resources to exceed certain threshold, further requests are throttled while the system allocates more resources using autoscaling.
`publish`	The `publish` API has a quota of 5000 requests per minute, per `appId`.

Artifact Service API quotas

API	Quota
`schema and artifact listing`	Each `appId` has a quota for requests that return lists. A single `appId` can send no more than 5 requests in 5 seconds
`download artifact with API key`	Each `appId` has a quota for requests that use an API key. A single `appId` can send no more than 500 requests in 5 seconds
`check access to an artifact`	Each `appId` has a quota for requests that return artifact. A single `appId` can send no more than 150 requests in 5 seconds
`other schema and artifact operations`	Each `appId` has a quota for requests that are not covered by previous paragraphs. A single `appId` can send no more than 50 requests in 5 seconds