Events and quotas are interconnected in Sentry. At the most basic, when you subscribe to Sentry, you pay for the number of events - errors, attachments, and transactions - to be tracked. Each of these type of events has a quota. When Sentry accepts an event, it counts toward your quota for that type of event.
Sentry’s flexibility means you can exercise fine-grained control over which events count toward your quota.
Let’s clarify a few terms to start:
- Event - an event is one instance of you sending Sentry data. Generally, this data is an error. Every event has a set of characteristics, called its fingerprint.
- Issue - an issue is a grouping of similar events, which all share the same fingerprint. For example, Sentry groups events together when they are triggered by the same part of your code. For more information, see Grouping & Fingerprinting.
- Attachment - Attachments are files uploaded in the same request, such as log files. Unless the option to store crash reports is enabled, Sentry will use the files only to create the event, and then drop the files.
- Transaction - A transaction represents a single instance of a service being called to support an operation you want to measure or track, like a page load.
- Quota - your quota is the monthly number of events - errors, attachments, and transactions - you pay Sentry to track.
Sentry completes a thorough evaluation of each event to determine if it counts toward your quota, as outlined in this overview. Detailed documentation for each evaluation is linked throughout. Before completing any of these evaluations, Sentry confirms that each event includes a valid DSN and project as well as whether the event can be parsed. In addition, for error events, Sentry validates that the event contains valid fingerprint information. If any of these items are missing or incorrect, the event is rejected.
SDK sample rate
If a sample rate is defined for the SDK, the SDK evaluates whether this event should be sent as a representative fraction of events. Setting a sample rate is documented for each SDK. The SDK sample rate is not dynamic; changing it requires re-deployment. In addition, setting an SDK sample rate limits visibility into the source of events. Setting a rate limit for your project may better suit your needs.
Events that exceed your quota are not accepted. Once your event volume is approaching or has exceeded the quota, teammates with the "owner" organization permission level will receive notification emails. To add to your quota or review what happens when you exceed it, see Increasing Quotas.
- If you have intervened to Delete and Discard an issue, then future events with the same fingerprint do not count toward your quota.
- If the previous event was resolved, this event counts toward your quota because it may represent a regression in your code.
- If you have intervened to ignore alerts about events with the same fingerprint, this event counts toward your quota because the event is still occurring. For more information, see Inbound Filters.
Sentry’s spike protection prevents huge overages from consuming your event capacity; spike protection is not currently available for transactions. For more information, see Spike Protection.
In addition, depending on your project’s configuration and the plan you subscribe to, Sentry may also check:
Rate limit for the project
If the error event rate limit for the project has been exceeded, and your subscription allows, the event will not be counted. For more information, see Rate Limiting in our guide to Manage Your Event Stream or Error Event Limits.
If any inbound filter is set for this type of error event, and your subscription allows, the event will not be counted. For more information, see Inbound Filters in our guide to Manage Your Event Stream or Inbound Data Filter.
After these checks are processed, the event counts toward your quota. It is accepted into Sentry, where it persists and is stored.
|Yes, this event counts||No, this event does not count|
|SDK configuration||Not set||Set to filter this event|
|SDK sample rate||Not set||Set|
|Are future error events that repeat set to Delete & Discard?||No|
|Is this repeated event resolved?||Yes, this may be a regression in your code|
|Is this repeated event set to Ignore Alert?||Yes, it is still occurring|
|Has spike protection been triggered?||No||Yes|
|Rate limit for error events for the project||Not set or not reached||Set and exceeded|
|Inbound Filters||Not set||Set for this error event|
Add to your quota at any time during your billing period, either by upgrading to a higher tier or by increasing your on-demand capacity. Our available plans suit most individual and business needs, and Sentry is designed to handle large throughput. If your team needs more, we’re happy to help. Reach out to our sales team at email@example.com to increase capacity.
If you’ve exceeded your quota threshold, the server will respond with a 429 HTTP status code. However, if this is your first time exceeding quota, you'll be entered into a one-time grace period. For more information, see this Help article. In addition, you can specify a spending cap for on-demand capacity if you need additional events; for example, if you’re rolling out a new version and anticipate more events this month. For more information, see On-Demand Spending Cap.
Per-key rate limits allow you to set the maximum volume of error events a key will accept during a period of time.
For example, you may have a project in production that generates a lot of noise. A rate limit allows you to set the maximum amount of data to “500 events per minute”. Additionally, you can create a second key for the same project for your staging environment, which is unlimited, ensuring your QA process is still untouched.
To set up rate limits, navigate to [Project] » Client Keys » Configure. Select an individual key or create a new one, then you’ll be able to define a rate limit as well as view a breakdown of events received by that key. For additional information and examples, see Rate Limiting in our guide to Manage Your Event Stream.
If you have enabled the storage of crash reports, you may set limits for the maximum number of crash reports that will be stored per issue. To set up these limits, use the slider in the "Store Native Crash Reports" option for your organization's Security and Privacy Settings.
In some cases, the data you’re receiving in Sentry is hard to filter, or you don’t have the ability to update the SDK’s configuration to apply the filters. Sentry provides several methods to filter data server-side, which apply before checking for potential rate limits. Inbound filters include:
- Common browser extension errors
- Events coming from localhost
- Known legacy browsers errors
- Known web crawlers
- By their error message
- From specific release versions of your code
- From certain IP addresses.
Explore these by navigating to [Project] » Project Settings » Inbound Filters. Commonly-set filters are discussed here for your quick reference. For additional information and examples, see Inbound Data Filters in our guide to Manage Your Event Stream. Commonly-set filters include:
If you have a rogue client, Sentry supports blocking an IP from sending data. Navigate to [Project] » Project Settings » Inbound Filters to add the IP addresses (or subnets) to Filter errors from these IP addresses.
If you discover a problematic release causing excessive noise, Sentry supports ignoring all events from that release. Navigate to [Project] » Project Settings » Inbound Filters, then add the releases to Filter errors from these releases.
Sentry supports filtering out a specific or certain kind of error as well. Navigate to [Project] » Project Settings » Inbound Filters, then add the error message to Filter errors by error message.
When you are unable to take immediate action on an issue, but it continues to occur, Sentry supports deleting and discarding that issue, which you can do from the Issue Details page. Navigate to the issue you would like to filter, click on the drop-down menu next to the bin icon, and select “Delete and discard future events”. This setting deletes most data associated with the issue and filters out new matching events before they count against your quota.
You can view and restore previously discarded issues by navigating to [Project] » Project Settings » Inbound Filters » Discarded Issues.
A spike is both a large and temporary increase in error event volume spike protection is not currently available for transactions. Sentry's spike protection prevents huge overages from consuming your error event capacity. We use your historical error event volume to implement a dynamic rate limit which discards error events when you hit its threshold. This dynamic rate limit is designed to protect you from short-term spikes. The threshold will increase if your increased volume persists for many hours.
When spike protection is activated, we limit the number of error events accepted in any minute to:
maximum(20, 6 x average error events per minute over the last 24 hours)
The 24 hour window ends at the beginning of the current hour, not at the current minute.
What this means is that if you experience a spike, we will temporarily protect you, but if the increase in volume is sustained, the spike protection limit will gradually increase until Sentry finally accepts all events.
Because Sentry bills on monthly event volume, spikes can consume your Sentry capacity for the rest of the month. If it really is a spike, spike protection drops error events during the spike to try and conserve your capacity. We also send teammates with the "owner" organization permission level a notification email.
Sentry imposes limits on various fields within an event, as well as the size of full events and the requests they are sent in:
- Events, attachments, and requests exceeding payload size limits are immediately dropped with a
413 Payload Too Largeerror. Sentry allows compressed content encoding, and applies separate limits before and after decompression.
- Fields exceeding the individual size limits are afterwards trimmed and truncated at a best effort.
To avoid unintentional data loss, consider limiting the size of values passed into Sentry's APIs. For example, if your application attaches application state or request bodies to Sentry events, truncate them first.
The precise limits may change over time. For more information, please refer to the following resources: