generated: '2026-08-23'
method: searched
source: https://console1.cloud.inspur.com/document/oss/1-service-introduction.html
docs:
- https://console1.cloud.inspur.com/document/oss/1-service-introduction.html
- https://console1.cloud.inspur.com/document/ecs/1-service-introduction.html
- https://console1.cloud.inspur.com/document/apig/1-service-introduction.html
note: 'Inspur Cloud publishes no API rate limits. Across all 16 documented API references
  there is no requests-per-second figure, no burst allowance, no 429 semantics, and no
  rate-limit response header of any kind — not X-RateLimit-*, not RateLimit-*, not
  Retry-After. An agent calling these APIs has no runtime signal telling it how close it is
  to being throttled, and no documented status code to recognise when it has been. What the
  provider does publish is RESOURCE QUOTAS — caps on how many objects of a type an account
  may hold — plus a traffic-control feature in the API Gateway PRODUCT that customers apply
  to their own published APIs, not to Inspur Cloud''s platform APIs. Both are recorded below
  and neither is a rate limit.'
limit_count: 0
response_headers: []
status_on_exhaustion: not documented
retry_after: not documented
rate_limits: []
resource_quotas:
- service: OSS
  scope: per account, per region
  limit: 200 buckets
  source: https://console1.cloud.inspur.com/document/oss/1-service-introduction.html
- service: OSS
  scope: per object, console upload
  limit: 5 GiB maximum single-file upload through the console
- service: OSS
  scope: per object, multipart upload
  limit: 48.8 TiB maximum object (max 5 GiB per part, max 10000 parts)
- service: OSS
  scope: per batch delete request, console
  limit: 1000 objects; larger batches must go through the API or SDK
- service: OSS
  scope: per bucket
  limit: 20 bound custom domains
- service: OSS
  scope: bandwidth
  limit: 'no per-user upload or download bandwidth limit is currently applied
    (单用户的上传和下载带宽暂时没有限制)'
- service: ECS
  scope: per availability zone
  limit: 'per-resource default quotas exist and differ by availability zone; the values are
    not published and are raised by contacting an account manager'
  source: https://console1.cloud.inspur.com/document/ecs/1-service-introduction.html
- service: ECS
  scope: per instance
  limit: security-group count per instance is capped; error 101.001018 is returned when the
    threshold is exceeded
- service: ECS
  scope: per region
  limit: private-image count per region is capped; error 101.003012 is returned when it is
    exceeded
- service: ECS
  scope: per request
  limit: batch key-pair deletion is capped at 100 keys (error 101.007011)
quota_exhaustion_signal:
  mechanism: error code in the response body
  codes:
  - code: '101.001021'
    meaning: user quota %s insufficient
  - code: '101.001020'
    meaning: failed to obtain user quota
  - code: '101.001018'
    meaning: adding security group failed, security-group count above threshold
  - code: '101.003012'
    meaning: private-image count in the current region may not exceed %s
  note: 'These are the closest thing to a throttling signal the platform publishes, and they
    are capacity errors rather than rate errors. They are returned in the body, not as a
    status code, so a client cannot distinguish them from any other failure without parsing.'
gateway_traffic_control:
  applies_to: customer-published APIs on the Inspur Cloud API Gateway product
  detail: 'API Gateway shipped 流量控制 (traffic control) in v1.5.0 (2020-04-24), refined it in
    v2.2.0 (2021-04-16) and added 熔断管理 (circuit-breaker management) in v2.5.0 (2021-12-14).
    These let an APIG customer impose limits on APIs they publish; they say nothing about the
    limits Inspur Cloud applies to its own platform APIs.'
  source: https://console1.cloud.inspur.com/document/apig/8-release-note.html