LP2053031: Adding tuning params

[FrancescoDeSimone]

this PR add tuning parameters like heartbeat_interval, election_timeout, snapshot_count, as describe in LP:#2053031

[reneradoi]

Hello @FrancescoDeSimone, I'm not sure if you saw the reply on Matrix, that's why I'll leave my comment here as well. Please refer here for further discussion.

[addyess]

Thank you so much for this PR and bringing attention to it. I hope to start a good conversation around this bug. So far the work looks good i only have a problem with the default for snapshot_count

[HomayoonAlimohammadi]

Great work, thanks a lot @FrancescoDeSimone! Left some comments and suggestions.

[addyess]

Only a tiny nit remains. Really nice work @FrancescoDeSimone

can you check out the unit test failures?

[addyess]

Beautiful work. LGTM 🍰

[addyess]

odd... the first run of the integration tests failed to complete test_snapshot_restore

• https://github.com/charmed-kubernetes/layer-etcd/actions/runs/12931100090/attempts/2?pr=213

I'll run again and see if it is successful

[addyess]

@FrancescoDeSimone well I guess it wasn't a fluke. The integration tests are failing in test_snapshot_restore due to these changes. DO you have time to look into why this failure occurs?

[FrancescoDeSimone]

hi @addyess try to understand what's the issue. Don't seems related to my changes. I tried to run the same test on the main branch and I got the same results

======================================== short test summary info =========================================
FAILED tests/integration/test_etcd.py::test_snapshot_restore - AssertionError: assert 'failed' == 'completed'
========================= 1 failed, 10 passed, 6 warnings in 1190.96s (0:19:50) ==========================
integration: exit 1 (1191.70 seconds) /home/ubuntu/test/layer-etcd> pytest -v -s --tb native --log-cli-level=INFO /home/ubuntu/test/layer-etcd/tests/integration pid=110510
  integration: FAIL code 1 (1204.01=setup[12.31]+cmd[1191.70] seconds)
  evaluation failed :( (1204.08 seconds)
ubuntu@ubuntu:~/test/layer-etcd$ git branch
* main

looks something relative to getent binary. I will try to dig deeper

unit-etcd-1: 18:48:14 ERROR unit.etcd/1.juju-log b'cmd_run.go:1276: WARNING: cannot create user data directory: cannot get the current user: getent could not be executed: exec: "getent": executable file not found in $PATH\ncmd_run.go:1281: WARNING: cannot copy user Xauthority file: cannot get the current user: getent could not be executed: exec: "getent": executable file not found in $PATH\nCouldn\'t find a member in the cluster with an ID of cmd_run.go.\n'

[FrancescoDeSimone]

@addyess I am currently facing the same problem with the upstream etcd, running this script

juju add-model test
juju deploy easyrsa 
juju deploy ch:etcd --to 0
juju relate etcd easyrsa
juju wait-for application etcd
juju run etcd/leader snapshot
juju scp etcd/leader:"/home/ubuntu/etcd-snapshots/etcd-snapshot-*.tar.gz" snapshot.tar.gz 
juju attach-resource etcd snapshot=./snapshot.tar.gz
juju run etcd/leader restore

got the same error in juju debug-log

unit-etcd-0: 19:29:27 ERROR unit.etcd/0.juju-log b'cmd_run.go:1276: WARNING: cannot create user data directory: cannot get the current user: getent could not be executed: exec: "getent": executable file not found in $PATH\ncmd_run.go:1281: WARNING: cannot copy user Xauthority file: cannot get the current user: getent could not be executed: exec: "getent": executable file not found in $PATH\nCouldn\'t find a member in the cluster with an ID of cmd_run.go.\n'

[addyess]

This will pass once the following is merged

• [LP#2096820] Don't push stderr through stdout when running etcdctl #215

[addyess]

@FrancescoDeSimone please rebase from main