Fix duplicate logging to Healthchecks and send "after_*" hooks output to Healthchecks (#328).

This commit is contained in:
Dan Helfman 2020-06-23 11:01:03 -07:00
parent 17c2d109e5
commit 37cc229749
5 changed files with 78 additions and 25 deletions

1
NEWS
View file

@ -1,5 +1,6 @@
1.5.7.dev0
* #327: Fix broken pass-through of BORG_* environment variables to Borg.
* #328: Fix duplicate logging to Healthchecks and send "after_*" hooks output to Healthchecks.
* #331: Add SSL support to PostgreSQL database configuration.
* #333: Fix for potential data loss (data not getting backed up) when borgmatic omitted configured
source directories in certain situations. Specifically, this occurred when two source directories

View file

@ -132,16 +132,6 @@ def run_configuration(config_filename, config, arguments):
if not encountered_error:
try:
if prune_create_or_check:
dispatch.call_hooks(
'ping_monitor',
hooks,
config_filename,
monitor.MONITOR_HOOK_NAMES,
monitor.State.FINISH,
monitoring_log_level,
global_arguments.dry_run,
)
if 'prune' in arguments:
command.execute_hook(
hooks.get('after_prune'),
@ -174,6 +164,24 @@ def run_configuration(config_filename, config, arguments):
'post-check',
global_arguments.dry_run,
)
if prune_create_or_check:
dispatch.call_hooks(
'ping_monitor',
hooks,
config_filename,
monitor.MONITOR_HOOK_NAMES,
monitor.State.FINISH,
monitoring_log_level,
global_arguments.dry_run,
)
dispatch.call_hooks(
'destroy_monitor',
hooks,
config_filename,
monitor.MONITOR_HOOK_NAMES,
monitoring_log_level,
global_arguments.dry_run,
)
except (OSError, CalledProcessError) as error:
if command.considered_soft_failure(config_filename, error):
return
@ -185,15 +193,6 @@ def run_configuration(config_filename, config, arguments):
if encountered_error and prune_create_or_check:
try:
dispatch.call_hooks(
'ping_monitor',
hooks,
config_filename,
monitor.MONITOR_HOOK_NAMES,
monitor.State.FAIL,
monitoring_log_level,
global_arguments.dry_run,
)
command.execute_hook(
hooks.get('on_error'),
hooks.get('umask'),
@ -204,6 +203,23 @@ def run_configuration(config_filename, config, arguments):
error=encountered_error,
output=getattr(encountered_error, 'output', ''),
)
dispatch.call_hooks(
'ping_monitor',
hooks,
config_filename,
monitor.MONITOR_HOOK_NAMES,
monitor.State.FAIL,
monitoring_log_level,
global_arguments.dry_run,
)
dispatch.call_hooks(
'destroy_monitor',
hooks,
config_filename,
monitor.MONITOR_HOOK_NAMES,
monitoring_log_level,
global_arguments.dry_run,
)
except (OSError, CalledProcessError) as error:
if command.considered_soft_failure(config_filename, error):
return

View file

@ -107,3 +107,15 @@ def ping_monitor(ping_url_or_uuid, config_filename, state, monitoring_log_level,
if not dry_run:
logging.getLogger('urllib3').setLevel(logging.ERROR)
requests.post(ping_url, data=payload.encode('utf-8'))
def destroy_monitor(ping_url_or_uuid, config_filename, monitoring_log_level, dry_run):
'''
Remove the monitor handler that was added to the root logger. This prevents the handler from
getting reused by other instances of this monitor.
'''
logger = logging.getLogger()
for handler in tuple(logger.handlers):
if isinstance(handler, Forgetful_buffering_handler):
logger.removeHandler(handler)

View file

@ -123,13 +123,13 @@ hooks</a> run, borgmatic lets Healthchecks know that it has started if any of
the `prune`, `create`, or `check` actions are run.
Then, if the actions complete successfully, borgmatic notifies Healthchecks of
the success before the `after_backup` hooks run, and includes borgmatic logs in
the success after the `after_backup` hooks run, and includes borgmatic logs in
the payload data sent to Healthchecks. This means that borgmatic logs show up
in the Healthchecks UI, although be aware that Healthchecks currently has a
10-kilobyte limit for the logs in each ping.
If an error occurs during any action or hook, borgmatic notifies Healthchecks
before the `on_error` hooks run, also tacking on logs including the error
after the `on_error` hooks run, also tacking on logs including the error
itself. But the logs are only included for errors that occur when a `prune`,
`create`, or `check` action is run.
@ -161,9 +161,9 @@ begins, ends, or errors. Specifically, after the <a
href="https://torsion.org/borgmatic/docs/how-to/add-preparation-and-cleanup-steps-to-backups/">`before_backup`
hooks</a> run, borgmatic lets Cronitor know that it has started if any of the
`prune`, `create`, or `check` actions are run. Then, if the actions complete
successfully, borgmatic notifies Cronitor of the success before the
successfully, borgmatic notifies Cronitor of the success after the
`after_backup` hooks run. And if an error occurs during any action or hook,
borgmatic notifies Cronitor before the `on_error` hooks run.
borgmatic notifies Cronitor after the `on_error` hooks run.
You can configure Cronitor to notify you by a [variety of
mechanisms](https://cronitor.io/docs/cron-job-notifications) when backups fail
@ -189,9 +189,9 @@ begins, ends, or errors. Specifically, after the <a
href="https://torsion.org/borgmatic/docs/how-to/add-preparation-and-cleanup-steps-to-backups/">`before_backup`
hooks</a> run, borgmatic lets Cronhub know that it has started if any of the
`prune`, `create`, or `check` actions are run. Then, if the actions complete
successfully, borgmatic notifies Cronhub of the success before the
successfully, borgmatic notifies Cronhub of the success after the
`after_backup` hooks run. And if an error occurs during any action or hook,
borgmatic notifies Cronhub before the `on_error` hooks run.
borgmatic notifies Cronhub after the `on_error` hooks run.
Note that even though you configure borgmatic with the "start" variant of the
ping URL, borgmatic substitutes the correct state into the URL when pinging

View file

@ -0,0 +1,24 @@
import logging
from flexmock import flexmock
from borgmatic.hooks import healthchecks as module
def test_destroy_monitor_removes_healthchecks_handler():
logger = logging.getLogger()
original_handlers = list(logger.handlers)
logger.addHandler(module.Forgetful_buffering_handler(byte_capacity=100, log_level=1))
module.destroy_monitor(flexmock(), flexmock(), flexmock(), flexmock())
assert logger.handlers == original_handlers
def test_destroy_monitor_without_healthchecks_handler_does_not_raise():
logger = logging.getLogger()
original_handlers = list(logger.handlers)
module.destroy_monitor(flexmock(), flexmock(), flexmock(), flexmock())
assert logger.handlers == original_handlers