pcdshub / ads-log-daemon Goto Github PK

Roadmap

1. Get log server online (request is in to PCDS IT)
2. For each PLC in plc-summary list (alternative method for getting all PLCs? use prog node staticroutes list?)
   a. Auto-route adding (borrow concept from pyads, or just use directly for now)
   b. Configure logging
   c. Save log messages to file and tail it with logstash OR send UDP messages to logstash directly

I think using the project name will be sufficient for most PLCs, but this may need to be revisited - or other experts may have an idea.

Initial implementation in #2 was exceedingly messy, attempting to get things rolling as quickly as possible.

This needs to be broken up and refactored to make it even remotely maintainable.

(It might be nice to use subprocesses at some point to keep the client instances separate, maybe... Perhaps can reconsider searching and parallel execution of tasks?)

Seems like all of the PLCs are similarly off - so there's some weird time zone interpretation happening (or not happening) here.

One or more of the following are probably off:

1. ads-async / ads-log-daemon's interpretation of the PLC timestamp
2. The timestamp shipped off to logstash
3. Logstash's interpretation of the timestamp in the JSON blob

Expected Behavior

• Provide more than one host for the LDAP server mode, and ads-log-daemon picks the first that works

Current Behavior

• Single only allowed
• LDAP host failure? No host updates

Current Behavior

• At some point, connections remain but no log messages are received
• Appears to require restart of ads-log-daemon, which I need to remember to do

Possible Solution

• Some keepalive request to the PLC?
• Reconnect after a certain period of time?
• Something else?