Crash pings (Telemetry) and crash reports (Socorro/Crash Stats): Part 2
Crash pings
Opt-out data set. Because of this, we think it's more representative of what's going on in the world--at least as much as Telemetry data.
It's a safe subset of crash annotations and the stack of the crashing process.
Nothing symbolicates that stack or computes crash signatures, so it's hard to glean anything from this data right now.
The data is located in the
telemetry.crashdata set. You can use Data Tools to access it like BigQuery, Redash, etc.
Crash reports
Opt-in data set. The user has to explicitly send the crash report. Because of this, we think it's not representative of what's going on in the world and that there are a lot of crashes that go unreported.
This data set is sampled--many incoming crash reports are rejected because this is category 4 data and we want to be holding as little as possible.
This contains all of the crash annotations as well as a lot of information from the crashed process including process memory. This is category 4 data!
This doesn't contain ids that can connect crash reports to other data sets on an individual level. For example, there are no Telemetry client ids in this data set allowing you to tie crash report data to Telemetry data.
Socorro symbolicates the stack and computes crash signatures, so we can look at Top Crasher reports and group crashes by signature.
Day |
# crash pings |
# crash reports |
|---|---|---|
2020-09-22 |
2,125,677 |
88,167 |
2020-09-23 |
2,060,468 |
90,203 |
2020-09-24 |
2,036,839 |
89,072 |
2020-09-25 |
2,159,750 |
87,587 |
2020-09-26 |
1,501,101 |
65,025 |
2020-09-27 |
1,575,065 |
64,199 |
2020-09-28 |
2,185,548 |
93,169 |