getsentry / getsentry/sentry

No org-level view of symbolication failures, and no bulk path to reprocess once symbols are fixed

Open
#122,956 3 comments 0 reactions 1 assignee View on GitHub

@shellmayr is already working on this.

Since Sep 1, 2026.

2. High Feature
Dominant language
Python
Stars
44.8k
Forks
4.9k
Avg merge
21h 23m
Merged PRs (30d)
607

Description

Problem Statement

When symbolication fails, the failure is only visible by opening an individual issue
and reading the Images Loaded section. There is no way to answer "which of my issues
have unsymbolicated or partially symbolicated frames", and therefore no way to know
how much of the crash dashboard is currently unreliable.

The recovery path has the same shape. Reprocessing is a per-issue button inside the
issue view. After uploading a missing or corrected debug file, there is no way to
reprocess everything that was blocked on it. For native teams that upload symbols
per build, a missed or late upload can affect thousands of issues, and each one has
to be found and reprocessed by hand.

The result is that:

  • Failures are discoverable only one issue at a time, so most are never discovered.
  • Teams have no measure of symbolication health, so a broken symbol upload in CI can
    go unnoticed for an entire release.
  • Fixing the cause does not fix the data, and the remediation cost scales with how
    long the problem went unnoticed.

This matters more than it might appear, because symbolication quality is upstream of
the stack trace, the grouping, and the fingerprint. Silent symbolication failure
degrades everything downstream of it.

Solution Brainstorm
  • A searchable, filterable view of symbolication outcomes across the org: issues and
    events with failed or partial symbolication, grouped by reason and by the debug
    file that was missing.
  • Search support so failed symbolication can be used as a filter in the issue stream.
  • Bulk reprocessing: select a set of issues, or all events blocked on a specific
    debug file, and reprocess in one action. Better still, automatically reprocess
    events that were blocked on a debug file when that file is later uploaded.
  • Alerting on symbolication failure rate crossing a threshold, so a broken CI upload
    is caught in hours rather than at the end of a release.

Related: #115602 (different root cause, same family of misleading candidate status)

Product Area

Processing

Requested by Saad Khan.

--

View Junior Session [Sentry]

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.