unable to initiate network scanning

[Mark Shuttleworth]

I see this in my regiond log every 5 minutes:

2018-05-06 16:19:25 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'scylla' (8tsb7m).
2018-05-06 16:19:25 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'maas' (4y3h7n).
2018-05-06 16:19:25 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'charybdis' (fbgtn8).
2018-05-06 16:19:25 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'lapsi' (4nqs36).
2018-05-06 16:19:25 maasserver.regiondservices.active_discovery: [info] Active network discovery: Unable to initiate network scanning on any rack controller. Verify that the rack controllers are started and have connected to the region.

Of these controllers, all but Lapsi report they are healthy.

[Matthew Davis (mdavistelstra)]

I don't think this is a duplicate of bug #1768575.

The string "ScanNetworks" doesn't even appear in that other bug report, and RPC is barely mentioned.

That thread is all about controllers being in a "degraded" state. My controller is marked as healthy, yet I still get the exact same problem as listed in this bug report.

[Alvin Cura (alvinc)]

I'm getting this too. And I have only one region+rack controller, so it's got nothing else to talk to.

2023-01-23 20:28:30 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'maas-temp' (fqpyps).
2023-01-23 20:28:30 maasserver.regiondservices.active_discovery: [info] Active network discovery: Unable to initiate network scanning on any rack controller. Verify that the rack controllers are started and have connected to the region.

[mitzone (mitzone)]

Hello,
I'm experiencing same issue with 3.2.7 and 3.3.1

2023-03-11 21:41:03 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'maas-test-rack1' (4cfb37).
2023-03-11 21:41:03 maasserver.regiondservices.active_discovery: [info] Active network discovery: Unable to initiate network scanning on any rack controller. Verify that the rack controllers are st
arted and have connected to the region.

Region and rack controller roles are separate.

Attaching logs for 3.2.7, fresh deb install.
Thanks.

[mitzone (mitzone)]

Same for a fresh install of 3.2.7 from snap.
Both roles on the same VM

2023-03-11 22:29:07 maasserver: [error] Error while calling ScanNetworks: Unable to get RPC connection for rack controller 'maas-test-region1' (yf47gy).
2023-03-11 22:29:07 maasserver.regiondservices.active_discovery: [info] Active network discovery: Unable to initiate network scanning on any rack controller. Verify that the rack controllers are started and have connected to the region.

[mitzone (mitzone)]

Hello,
I have this behaviour on a fresh install of 3.3.x latest and 3.2.7 for both DEB and SNAP installs.
Region and rack on same box. ScanNetworks not working.
Thank you.

[Igor Brovtsin (igor-brovtsin)]

For some reason, the logs ~mitzone provided for the rack controller have some gaps. "Unable to get RPC connection for rack" errors timing matches these gaps:

- 21:35:58 -> 21:38:35 no logs for rack, 21:36:03 region logs "Unable to get RPC connection"
- 21:39:00 rack logs end, 21:41:03 and 21:46:03 region logs "Unable to get RPC connection"

Judging by the RPC connection registration logs, the time is in sync for region and rack.

[mitzone (mitzone)]

Hello,
Those are not gaps. There is no activity. That is a test env.
Issue was already reproduced and confirmed here: https://bugs.launchpad.net/maas/+bug/1953049
Cheers.

[Igor Brovtsin (igor-brovtsin)]

Dup since the thread ~mitzone linked contains more information relevant to the current MAAS version.