mirror of
https://github.com/hanzoai/traces.git
synced 2026-10-30 18:20:41 +00:00
76c2c85b97
* [Generator] Add DRAIN span_name sanitizer * Fix linter * Fix formatting * Update default manifest * Fix build error in update-fixtures * Remove bad test * Remove misleading comment * Add per-tenant override for drain mode * Do a machines job * Address PR feedback * Address PR review: rename drain_mode config and improve sanitizer - Rename `drain_mode` to `span_name_sanitization` across all config, legacy config, interfaces, and overrides for better user-facing clarity - Add explicit default value "disabled" in RegisterFlagsAndApplyDefaults - Add ValidateSpanNameSanitization function in validation/fields.go - Rename metrics to remove DRAIN reference: `drain_spans_compacted_total` -> `spans_sanitized_total` `post_drain_demand_estimate` -> `post_sanitization_demand_estimate` - Pass configured staleDuration to NewDrainSanitizer instead of hardcoded 15min, consistent with other registry components - Extract "span_name" magic string to labelSpanName constant - Update all tests with new names and signatures * Update manifest * Use empty string as disabled * Update CHANGELOG.md Co-authored-by: Suraj Nath <[email protected]> * Update manifest --------- Co-authored-by: Suraj Nath <[email protected]>
58 lines
1.2 KiB
Go
58 lines
1.2 KiB
Go
package drain
|
|
|
|
type LogCluster struct {
|
|
id int
|
|
Size int
|
|
Tokens []string
|
|
Stringer func([]string) string
|
|
ParamString string
|
|
cache string
|
|
}
|
|
|
|
func (c *LogCluster) ingestTokens(tokens []string) {
|
|
if len(tokens) != len(c.Tokens) {
|
|
panic("attempt to create template from sequences with different token lengths")
|
|
}
|
|
for i := range tokens {
|
|
if tokens[i] != c.Tokens[i] && c.Tokens[i] != c.ParamString {
|
|
c.Tokens[i] = c.ParamString
|
|
c.cache = ""
|
|
}
|
|
}
|
|
c.Size++
|
|
}
|
|
|
|
func (c *LogCluster) tokenDistance(tokens []string) (float64, int) {
|
|
if len(c.Tokens) != len(tokens) {
|
|
panic("attempt to compare sequences with different token lengths")
|
|
}
|
|
|
|
similarTokens := 0
|
|
paramCount := 0
|
|
for i := range c.Tokens {
|
|
clusterToken := c.Tokens[i]
|
|
inputToken := tokens[i]
|
|
|
|
switch clusterToken {
|
|
case c.ParamString:
|
|
paramCount++
|
|
case inputToken:
|
|
similarTokens++
|
|
}
|
|
}
|
|
retVal := float64(similarTokens) / float64(len(c.Tokens))
|
|
return retVal, paramCount
|
|
}
|
|
|
|
func (c *LogCluster) String() string {
|
|
if c.cache != "" {
|
|
return c.cache
|
|
}
|
|
c.cache = c.Stringer(c.Tokens)
|
|
return c.cache
|
|
}
|
|
|
|
func (c *LogCluster) GetTokens() []string {
|
|
return c.Tokens
|
|
}
|