1951450c57
Usage: from a host for the affected satellite, issue this command
and supply the number of segments that have been permanently lost. For
example, to indicate 2 segments permanently lost:
satellite register-lost-segments 2
If the unthinkable happens and this is necessary to use, it is presumed
that we will also remove the non-recoverable segments from the metainfo
db, so they will not continue to appear as 'temporarily unavailable' to
the repair checker.
The influxQL query for this will probably look something like:
SELECT sum(total) AS "sum_total"
FROM "v3_stats_new"."autogen"."lost-segments"
WHERE time > :dashboardTime: AND "scope" = 'segment-durability'
GROUP BY time(:interval:), "application", "instance"
FILL(null)
Or use the Flux language in order to get the benefit of the
cumulativeSum function:
from(bucket: "v3_stats_new/autogen")
|> range(start: dashboardTime)
|> filter(fn: (r) => r._measurement == "lost-segments"
and r.scope == "segment-durability"
and r._field == "total")
|> group(columns: ["application", "instance"])
|> cumulativeSum()
Change-Id: I73d364937705aa815af7520ab79c00bc2aea09f6
|
||
|---|---|---|
| .. | ||
| certificates | ||
| connect-test | ||
| crashcollect | ||
| identity | ||
| inspector | ||
| internal/wizard | ||
| metabase-orphaned-segments | ||
| metabase-verify | ||
| metric-receiver | ||
| multinode | ||
| satellite | ||
| storagenode | ||
| storagenode-updater | ||
| storj-admin | ||
| storj-sim | ||
| uplink | ||
| uplinkng | ||
| versioncontrol | ||