Ingestion limits

The default ingestion limits in Tempo may not be sufficient in high volume tracing environments. The following message, for example, indicates that a tenant is exceeding the trace limits imposed on them.

   max live traces per tenant exceeded: per-user traces limit (local: 10000 global: 0 actual local: 10000) exceeded

The following sections describe the available options, per tenant overrides and global vs. local configurations.

Configuration options

The following options can be used to limit ingestion:

• ingestion_burst_size_bytes : Burst size (bytes) used in ingestion. Default is 20,000,000 (~20MB).
• ingestion_rate_limit_bytes : Per-user ingestion rate limit (bytes) used in ingestion. Default is 15,000,000 (~15MB).
• max_bytes_per_trace : Maximum size of a single trace in bytes. 0 to disable. Default is 5,000,000 (~5MB).
• max_traces_per_user: Maximum number of active traces per user, per ingester. 0 to disable. Default is 10,000.

Both the ingestion_burst_size_bytes and ingestion_rate_limit_bytes parameters control the rate limit. When these limits exceed the following message is logged:

    RATE_LIMITED: ingestion rate limit (15000000 bytes) exceeded while adding 10 bytes

When the limit for the max_bytes_per_trace parameter exceeds the following message is logged:

    TRACE_TOO_LARGE: max size of trace (5000000) exceeded while adding 387 bytes

Finally, when the limit for the max_traces_per_user parameter exceeds the following message is logged:

LIVE_TRACES_EXCEEDED: max live traces per tenant exceeded: per-user traces limit (local: 10000 global: 0 actual local: 1) exceeded

Standard overrides

To configure new ingestion limits that applies to all tenants of the cluster:

1. Create an overrides section at the bottom of your configuration file.
2. Add the parameters under this section. For example:

    overrides:
        ingestion_burst_size_bytes: 5_000_000
        ingestion_rate_limit_bytes: 5_000_000
        max_bytes_per_trace: 5_000_000
        max_traces_per_user: 10000 

A snippet of a config.yaml file showing how the overrides section is here.

Tenant-specific overrides

Sometimes you don’t want all tenants within the cluster to have the same settings. To add overrides that are specific to an individual tenant:

1. Create an overrides section at the bottom of the config.yaml file.

2. Add an entry per_tenant_override_config to point to a separate file named overrides.yaml that contains tenant-specific overrides.

   overrides:
       per_tenant_override_config: /conf/overrides.yaml
   

3. Within the overrides.yaml file, add the tenant-specific override parameters. For example:

   overrides:
       "<tenant id>":
           ingestion_burst_size_bytes: 5_000_000
           ingestion_rate_limit_bytes: 40_000_000
           max_bytes_per_trace: 25_000_000
           max_traces_per_user: 100000
   

This overrides file is dynamically loaded. It can be changed at runtime and will be reloaded by Tempo without restarting the application.

1. Additionally a “wildcard” override can be used that will apply to all tenants if a match is not found otherwise.

    overrides:
        "*":
            ingestion_burst_size_bytes: 5_000_000
            ingestion_rate_limit_bytes: 40_000_000
            max_bytes_per_trace: 25_000_000
            max_traces_per_user: 100000

Override strategies

The trace limits specified by the various parameters are, by default, applied as per-distributor limits. For example, a max_traces_per_user setting of 10000 means that each distributor within the cluster has a limit of 10000 traces per user. This is known as a local strategy in that the specified trace limits are local to each distributor.

A setting that applies at a local level is quite helpful in ensuring that each distributor independently can process traces up to the limit without affecting the tracing limits on other distributors.

However, as a cluster grows quite large, this can lead to quite a large quantity of traces. An alternative strategy may be to set a global trace limit that establishes a total budget of all traces across all distributors in the cluster. The global limit is averaged across all distributors by using the distributor ring.

The default setting for ingestion strategy is local. The strategy setting can also be changed using the overrides section, as shown in the following example:

overrides:
    ingestion_rate_strategy: global