<signals>
Data for export accumulates as files in a separate directory — /var/lib/pmta/sparkpost on Unix — and is deleted after a successful upload. The line sets the free space threshold below which the node starts deleting them without sending. The default is 20M.
🔴 What is deleted is lost forever, and the service’s reports are left with gaps. The vendor says so directly.
In normal operation the threshold does not fire: files are sent and deleted on their own. They start to accumulate only when the service is unavailable, overloaded or misconfigured.
So this is not a setting but a safety fuse. It decides what to sacrifice during a prolonged outage: disk space or analytics data.
The choice here is short:
| Value | When |
|---|---|
| 20M (default) | a separate partition for export |
| 1G and more | a partition shared with the spool and accounting |
| 0 | never: the disk will run out completely |
# the export directory is on a shared partition — raise the threshold <signals> api-key 0123456789abcdef012345678901234567890123 min-free-space 1G</signals>
what people get wrong and how it ends1 mistake▶
our experience
2 relations
du -sh /var/lib/pmta/sparkpost 2>/dev/null; df -h /var/lib/pmta
3 articles
◇What the node tells about itself and through which channelsSeven observation surfaces: the web monitor, the console in four formats, the log, accounting, accounting search, metrics and export to external analytics.◇Two inbound report processors and what to do with themBounces by message and complaints are parsed by two identically structured blocks; each has its own address list and its own pair of decisions about the message’s fate.◇Five ways for a queue to hand off a messageA queue does not have to use SMTP: it can write to a file, run a program, call over HTTP or discard.