nethvoice-alerts service fails too much generating alerts sometime

[Stell0]

if nethvoice-alerts service fails to communicate with Asterisk, it exit with error and systemd restart it. But sometimes this generate an alert on my.nethesis.it because service status is checked every 5 minutes and nethvoice-alerts take 30 second before respawn, therefore service status could be checked during the 30 seconds downtime.

Steps to reproduce

• send a SIGKILL to asterisk
• nethvoice-alerts connection should fail

Expected behavior

• nethvoice-alerts should just reconnect and keep going

Actual behavior

• service fails and is restarted after 30 seconds by systemd

[gsanchietti]

Also recallonbusy service is affected by the same problem.

[nethbot]

in 7.9.2009/nethesis-testing:

• nethserver-nethvoice14-14.16.3-1.1.g208c97c.ns7.noarch.rpm x86_64

[nethbot]

in 7.9.2009/nethesis-testing:

• recallonbusy-1.0.2-1.2.g11aa24d.ns7.noarch.rpm x86_64

[gsanchietti]

Test case

• Make sure recallonbusy and nethvoice-alerts are up&running
• Stop asterisk daemon
• Verify both services do not die
• Restart asterisk daemon
• Verify both services reconnect correctly to asterisk

[nethbot]

in 7.9.2009/nethesis-testing:

• recallonbusy-1.0.2-1.3.g40fe70f.ns7.noarch.rpm x86_64

[gsanchietti]

The problem seems fixed on both daemons. We are going to keep it running for a while to exclude regressions.

[gsanchietti]

Verified.

[nethbot]

in 7.9.2009/nethesis-updates:

• recallonbusy-1.0.3-1.ns7.noarch.rpm x86_64

[nethbot]

in 7.9.2009/nethesis-updates:

• nethserver-nethvoice14-14.16.4-1.ns7.noarch.rpm x86_64