Cantal Changes by Version¶
Cantal 0.6.10¶
- Bugfix: previously swap metric wasn’t seen in metrics (only in all_processes) and 0 was always submitted to graphite/carbon
Cantal 0.6.9¶
- Bugfix: previously cantal sometimes skipped some processes because scanning /proc is not atomic. Now we decrease the issue by making scan fast and doing it twice.
Cantal 0.6.8¶
- Bugfix: machine uptime and process’ uptimes were broken in UI
Cantal 0.6.7¶
- Bugfix: previously cantal could crash when time jumps backwards. Currently, it waits (in metrics scanner code) if delta is < 10 sec and crashes with clear log message for large time jumps.
Cantal 0.6.6¶
- Bugfix: when a
CANTAL_PATH
referring to a same file is encountered in multiple processes we no longer duplicate metrics
Cantal 0.6.5¶
- Feature: add
peers
graphql field
Cantal 0.6.4¶
- Feature: add
local.cgroups
graphql endpoint - Feature: add
local.processes
graphql endpoint
Cantal 0.6.3¶
- Bugfix: add
num_peers
,num_stale
back to/status.json
, same fields added to graphql endpoint
Cantal 0.6.2¶
- Bugfix: larger timeouts for incoming http requests
- Bugfix: add
version
back to/status.json
Cantal 0.6.1¶
- Bugfix: fix JS error on /local/peers page
Cantal 0.6.0¶
- We reworked network subsystem to use tokio instead of home-grown async, this looses some features for now, but is an important step for future
- Breaking: remote subsystem doesn’t work, including the whole
/remote
route, we will be working to add feature back soon - Feature: add graphql API (only status for now)
- Breaking:
/status.json
contains less data, use graphql API