Limitations

Rate limiting

Rate limiting is applied to each API key and control the number of calls or messages per time window (e.g. 1 call per second). Depending on the offer, a rate limiting configuration may be applied to http interface, mqtt interface or both.

Http interface

Each response of the web controller contains 3 headers giving additional information on the status of the current request regarding rate limitation:

X-RateLimit-Limit: 5
X-RateLimit-Remaining: 3
X-RateLimit-Reset: 1479745936295
  • X-Rate-Limit-Limit is the rate limit ceiling per second

  • X-Rate-Limit-Remaining is the number of requests left for this time window

  • X-RateLimit-Reset is the ending date of the current time window (expressed in epoch milliseconds).

When receiving a request that would exceed authorized traffic limit, the web application returns a 429 Too Many Requests error with an empty body.

Note that all X-RateLimit headers are present in the response, as they would in a successful response.

Mqtt limits

Live Objects enforces limits for connections and traffic. Refer to the limit table. When a MQTT client tries to connect to Live Objects, connection limits are enforced according to the connection mode ( Device, Application, Connector ). If a limit is reached, the connection is refused. When a MQTT client sends traffic to Live Objects ( publish, subscribe ), traffic limits are enforced according to the connection mode ( Device, Application, Connector ). Refer to the following section for trafic limit policies.

Mqtt trafic limits with Device Mode

Devices using MQTT are subject to the subscribed offer traffic limits. The limits are at the account level and at the connection level. Refer to the limit table for details.

If a limit is reached for a MQTT/publish then the published message will be dropped and a warning will be available in Audit Log.

If a limit is reached for a MQTT/subscribe then the subscribe message will be rejected and a warning will be available in Audit Log.

Mqtt trafic limits with Application and Connector Mode

If a traffic ( subscribe, publish ) limit is reached, then the MQTT session will be closed. If an API key is used for several MQTT sessions at the same time, the API key quotas are computed with the overall API key traffic.

As MQTT 3.11 protocol does not provide a way to give disconnection reasons; you should use the AuditLog service to retrieve additional information. The client is expected to reconnect with a back off policy and re-send its data, as traffic will be allowed again in the next time window.

Resources limitation

Resource rate limiting is applied to each API Key and limit the search request.

X-ResourceLimit-Limit : 100
X-ResourceLimit-Remaining : 10
X-ResourceLimit-Reset : 1568707112000
X-ResourceLimit-Resource :SEARCH_DSL_QUERY
  • X-ResourceLimit-Limit is the rate limit ceiling of the authorized traffic limit (number of requests) for this time window

  • X-ResourceLimit-Remaining is the number of requests left for this time window

  • X-ResourceLimit-Reset is the ending date of the current time window (expressed in epoch milliseconds).

  • X-ResourceLimit-Resource is the resource identifier (SEARCH_DSL_QUERY)

When receiving a request that would exceed authorized traffic limit, the web application returns a 429 Too Many Requests error with an empty body.

Limitation table

Category Limit Limit type Value Default Auditlog notification

Access Management

Max number of API-Key

Absolute

Check your offer

5

NO

Max number of users

Absolute

Check your offer

5

NO

Triggers & Actions : Triggers

Max number of connectors filter per filter’s type

Absolute

Non-adjustable

20

NO

Triggers & Actions : Actions

Max number of action policy

Absolute

Check your offer

20

NO

Max number of each action’s type per action policy

Absolute

Check your offer

3

NO

Max number of emitted emails

Rate

Check your offer

1000/day

YES

Max number of emitted SMS

Rate

Check your offer

10/day

NO

(httpPush) Max waiting time for a response from the remote host

Rate

Non-adjustable

5 seconds

YES

(httpPush) Socket timeout: Max time of inactivity between two data packets after connection

Rate

Non-adjustable

5 seconds

YES

(httpPush) Max number of bufferized messages (overflow)

Absolute

Non-adjustable

10.000

YES

Triggers & Actions : Tests

HttpPush message tests (/api/v1/event2action/test/http-push)

Rate

Non-adjustable

5 per minutes

NO

FIFO message tests (/api/v1/event2action/test/fifo)

Rate

Non-adjustable

5 per minutes

NO

Alarming

Max Number of Matching rules

Absolute

Check your offer

50

NO

Max Number of Firing rules

Absolute

Check your offer

50

NO

Max Number of State processing rules

Absolute

Check your offer

50

NO

Max Number of Activity processing rules

Absolute

Check your offer

50

NO

Max number of deviceIds per Activity processing rule

Absolute

Non-adjustable

100

NO

Max number of groupPaths per Activity processing rule

Absolute

Non-adjustable

100

NO

Auditlog

Max number of audit log messages. LoRa logs are not included in this quota computation.

Absolute

Non-adjustable

10.000.000

NO

Log messages retention delay

Absolute

Non-adjustable

14 days

YES

Business application & REST API

Max number of HTTP requests per API-Key

Rate

Check your dashboard

120/minute

NO

Max number of HTTP requests per account

Rate

Check your dashboard

3600/hour

NO

Max number of HTTP concurrent requests per account (depending on the invoked end-point)

Rate

Non-adjustable

5-80 connections

NO

Max number of emitted bulks (for more information, see the chapter "Add a data message to a stream")

Rate

Check your dashboard

120/minute

NO

Max number of messages per bulk

Absolute

Check your offer

1000

NO

Max number of created FIFO (for more information, see the chapter "Using FIFO")

Absolute

Check your offer

10

NO

FIFO retention. Maximum is 7 days.

Absolute

Check your offer

7 days

NO

FIFO list forceRefresh rate-limits.

Absolute

Non-adjustable

total of 20 seconds of compute / 60 seconds window (★)

NO

MQTT: Application mode

Max number of MQTT concurrent connections using application mode

Absolute

Check your dashboard

100

YES

Max connection attempts per account using application mode

rate

Non-adjustable

1/second

YES

Max number of MQTT message published per API-Key using: application mode

Rate

Check your dashboard

120/minute

YES

Max number of MQTT message published in total for the account using: application mode

Rate

Check your dashboard

3600/hour

YES

MQTT: Device mode

Max number of MQTT concurrent connections using device mode

Absolute

Check your offer

100

YES

Max connection attempts per device using device mode

Rate

Non-adjustable

1/second

YES

Max connection attempts per account using device mode

Rate

Check your offer

5/second

YES

Max number of MQTT message published per connection using: device mode

Rate

Check your offer

120/minute

YES

Max number of MQTT message published in total for the account using: device mode

Rate

Check your offer

3600/hour

YES

Max number of MQTT subscribe per connection using: device mode

Rate

Non-adjustable

10/minute

YES

MQTT: External connector mode

Max number of MQTT concurrent connections using external connector mode

Absolute

Check your dashboard

100

YES

Max connection attempts per account using external connector mode

rate

Non-adjustable

1/second

YES

Max number of MQTT message published per API-Key using: external connector mode

Rate

Check your dashboard

120/minute

YES

Max number of MQTT message published in total for the account using: external connector mode

Rate

Check your dashboard

3600/hour

YES

Data Management

Max number of custom pipelines

Absolute

Check your offer

10

NO

Max number of bufferized messages (overflow) for externalTransformation step

Absolute

Non-adjustable

10.000

NO

http connection timeout for externalTransformation step

Absolute

Non-adjustable

15 seconds

NO

http socket timeout for externalTransformation step

Absolute

Non-adjustable

30 seconds

NO

Max number of data retrieved per request using /streams http API (streamId)

Absolute

Non-adjustable

1.000

NO

Max number of data retrieved using /streams http API (streamId)

Rate

Check your offer

25.000/minute

NO

Data messages retention delay in store service

Absolute

Check your offer

1 year

NO

Data Messages size limits

Max size of useful payload (size of "value" field).

Absolute

Non-adjustable

1 MBytes

NO

Max number of tags in datamessage

Absolute

Non-adjustable

10

YES

Device Management: Inventory

Max number of devices

Absolute

Check your offer

10

NO

Max number of tags

Absolute

Check your offer

10/device

NO

Max number of properties

Absolute

Check your offer

10/device

NO

Max number of interfaces

Absolute

Non-adjustable

5/device

NO

Max number of groups

Absolute

Non-adjustable

1500

NO

Max number of group levels

Absolute

Non-adjustable

7

NO

Max duration of query processing time using /stats http API

Rate

Non-adjustable

1 second of query processing time per 10 seconds

NO

Max number of data retrieved using /search http API

Rate

Non-adjustable

10.000 devices per 5 seconds

NO

Device Management: Resources

Max size of a resource file

Absolute

Non-adjustable

10 MBytes

NO

Max number of aliases per resource

Absolute

Non-adjustable

5

NO

Size of device error message can be published: Max number of chars in message

Absolute

Check your offer

256

YES

Max number of resource per device

Absolute

Non-adjustable

50

YES

In-band update max number of retries for READ Firmware state operation

Absolute

Non-adjustable

10

NO

In-band update max number of retries for READ Firmware update result operation

Absolute

Non-adjustable

20

NO

Device Management: Command

Max number of queued commands per device

Absolute

Non-adjustable

10

NO

Device Management: Campaigns

Max number of scheduled campaigns: number of ongoing + scheduled excluding the completed and removed campaigns.

Absolute

Check your offer

50

NO

Max number of devices target in a campaign

Absolute

Check your offer

200

NO

Max number of operations in a campaign

Absolute

Non-adjustable

5

NO

Twin Management: Model

Max number of custom models

Absolute

Non-adjustable

20

NO

Twin Management: Operations

Max number of actives operations

Absolute

Check your offer

30

NO

Max number of final operations

Absolute

Check your offer

100

NO

Max number of attempts

Absolute

Non-adjustable

5

NO

Max global size of values for a write operation

Absolute

Non-adjustable

1Mo

NO

Twin Management: Reported values

Max distinct path for reported value per device

Absolute

Non-adjustable

500

YES

Max global size of reported values per device

Absolute

Non-adjustable

1Mo

YES

Max data message size of reported values

Absolute

Non-adjustable

1Mo

YES

Twin Management: Observation

Max number of observations per device

Absolute

Non-adjustable

10

NO

LWM2M connector

Size of a resource body that will be accepted as the payload of a CoAP POST/PUT or the response to a GET request

Absolute

Non-adjustable

550 KBytes

NO

Max number of supported and instantiated objects in the Register and Update operations

Absolute

Non-adjustable

200

YES

Max number of uplinks linked to the Registration Interface (Register / Update / De-register) for a given endpoint

Rate

Non-adjustable

10/minute

YES

Max number of uplinks linked to the Information Reporting Interface (Notify / Send) for a given endpoint

Rate

Non-adjustable

60/minute

YES

SMS connector

Max number of SMS interfaces

Absolute

Check your offer

10

NO

Max number of outgoing messages

Rate

Check your offer

100/day

NO

LoRa connector

Max number of LoRa interfaces

Absolute

Check your offer

10

NO

Max number of outgoing messages

Rate

Check your offer

None

NO

Decoding service

Max number of CSV decoders

Absolute

Non-adjustable

100

NO

Max number of binary decoders

Absolute

Non-adjustable

100

NO

Max number of javascript decoders

Absolute

Non-adjustable

100

NO

Max number of javascript-split decoders

Absolute

Non-adjustable

100

NO

Bootstrap LwM2M service

Max duration of the compute time during requests in the http /bootstrap/lwm2m/entries/search and /bootstrap/lwm2m/entries/count API. For each request, the compute time will be taken into account and the sum should not exceed the quota

Rate

Non-adjustable

total of 1 sec of compute / 10 sec

NO

Indexation & Search

Data messages retention delay in search service

Absolute

Check your offer

1 year

NO

Max number of indexed fields for the account in total

Absolute

Check your offer

1000

YES

Max duration of the compute time during requests in the http search API. For each request, the compute time will be taken into accountand the sum should not exceed the max

Rate

Check your offer

total of 10 sec of compute / 10 sec

NO

Security

Max number of CA certificates for SSL2WAYS

Absolute

Check your offer

10

NO

Network metrics

Max retention of network metrics

Absolute

Non-adjustable

31 days

NO

(★) : some limits are related to LiveObjects platform live load. Client must be able to manage 429 Too Many requests response on his own by retrying the request.