Project

General

Profile

Bug #23344

Node down errors from aliased acnet nodes since erlang dpm shutdown

Added by Dennis Nicklaus about 2 months ago. Updated about 2 months ago.

Status:
Assigned
Priority:
Normal
Assignee:
Category:
Data Pool Manager
Target version:
-
Start date:
09/27/2019
Due date:
% Done:

0%

Estimated time:
Duration:

Description

Probably since the Erlang DPMs were shut down, ACNET returns [1 -42] (ACNET_NODE_DOWN) errors for devices from the aliased acnet nodes for CLX58. There are several (CMLRFA, CMLRFB, CMLRFC, CMLRFD, and STCLRF) nodes which are aliased to CLX58 (CLX58E?). What had previously worked was for the Erlang front-end to just run as the default node CLX58, not the normal CLX58E. That let things work. Now, we get 1 -42 errors both from the aliased node devices (e.g. T:1LQID, T:SQSYID) and from devices which are on CLX58E itself (e.g. U:GRWD8A, T:CDLPCC).

When we run the Erlang front-end as default node CLX58E, the devices assigned to it (U:GRWD8a) work, but the others (aliases) still don't ([1 -42] errors).

Rich suggests that having Charlie restart the consolidating java node will clear this. But that seems like a non-ideal temporary solution.

History

#1 Updated by Richard Neswold about 2 months ago

  • Status changed from New to Assigned
  • Description updated (diff)

Just a comment to add: the forwarding-requests-to-the-default-node feature was a hack to get OACs to work with virtual nodes. This feature wasn't advertised because it was never intended to be permanent. Charlie has been slowly moving the Java world over to acnetd and when that's complete, this feature was to go away.

The better, long-term solution is to break your front-end into five. The mk_acsysfe script would make this easy to do.

#2 Updated by Denise Finstrom about 2 months ago

A temporary fix was made to the acnet startup script to restore the default node fallback behavior on clx58.



Also available in: Atom PDF