MAX_IDS
MAX_IDS = 32
Commit ids one bucket carries.
What happened inside one bucket of a timeline.
Holds the commit ids as well as the totals, because a bucket is clickable: a reader who sees a spike wants the commits that made it, and re-querying by time range would give a different set from the one that was drawn if a flush landed in between.
The id list is bounded. A bucket at the day resolution on a busy site covers thousands of commits, and a chart holding all of them would spend more on the ids than on the numbers. Past the cap the count still rises, so the bar's height stays honest and only the drilldown is partial.
__construct(int $startMicrotime, int $endMicrotime, int $commits = 0, int $operations = 0, int $rawBytes = 0, int $storedBytes = 0, int $anchors = 0, list$ids = [], int $restores = 0) : mixed
Constructs a bucket.
| int | $startMicrotime | The bucket's start, in unix microseconds. |
| int | $endMicrotime | The bucket's end, in unix microseconds. |
| int | $commits | How many commits fell in it. |
| int | $operations | Operations those commits carried. |
| int | $rawBytes | Bytes before compression. |
| int | $storedBytes | Bytes actually written. |
| int | $anchors | How many of the commits were base anchors. |
| list |
$ids | Commit ids, up to MAX_IDS of them. |
| int | $restores | How many restores were run inside the bucket. |