Manage pipes - Pipedata

pd ls shows every pipe in your workspace with its state and stats. pause / resume toggle whether a pipe accepts new data or delivers records. delete requests removal.

TL;DR

$ pd ls                       # all pipes, all-time stats
$ pd ls --range 1h            # rates over the last hour
$ pd pause  events            # stop accepting / delivering
$ pd resume events            # start again
$ pd delete events            # interactive confirm; pass --yes to skip

`pd ls`

$ pd ls
NAME       KIND         FROM    STATUS   REC      BACKLOG  NOTES
orders     source:http  —       running  12,304   0        —
clean      uniq         orders  running  12,180   0        —
webhook    dest:http    clean   running  12,180   0        —
old-test   source:http  —       paused   500      0        —

--range controls the window for stats (15m, 1h, 24h, 7d, 30d, all). Default is all. The NOTES column shows transient delivery failures (* fails, retried automatically) and permanent undelivered counts (records that exceeded the retry budget). See Forward to a webhook for retry behavior.

Pipe states

Status	Meaning
`running`	Reading / accepting / delivering.
`paused`	Stopped. Producers can’t push to a paused source; downstream pipes pause too.

`pd pause` / `pd resume`

Both take a pipe name and toggle its running flag. Works for any pipe kind (source:http, dest:http, uniq).

$ pd pause  events
$ pd resume events

Pausing a source pauses everything downstream too — pipes that depend on it. Resume the source first, then the downstream pipes.

`pd delete`

$ pd delete events           # prompts: type the pipe name to confirm
$ pd delete events --yes     # skip the prompt (scripts, CI)

Deletion is irreversible. Records are purged within 72 hours and cannot be recovered. Pipes downstream of the deleted pipe are paused.

​TL;DR

​pd ls

​Pipe states

​pd pause / pd resume

​pd delete

​See also

TL;DR

`pd ls`

Pipe states

`pd pause` / `pd resume`

`pd delete`

See also