[INSTRM-563] Finalize enu hardware architecture Created: 05/Nov/18  Updated: 22/Nov/19  Resolved: 22/Nov/19

Status: Won't Fix
Project: Instrument control development
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Normal
Reporter: arnaud.lefur Assignee: Yoshida, Hiroshige
Resolution: Won't Fix Votes: 0
Labels: SM1, SPS, subaru-personnel
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Relates
relates to INSTRM-806 Provision rack5-ics host Done
relates to INSTRM-803 Install enuActor and all sps actors a... Done
Story Points: 4
Sprint: SM1-2019 L, SM1-2019 N, SM1PD-2019 B

 Description   

Since we have decided to get rid of PC104 and use a moxa for serial devices, enuActor(s) can be run on any machine which is on the PFS network.

There are several options, but a choice has to be made at some point.

We can run 4 enuActor instances on one machine :

  • either on a dedicated server/RasPI
  • an existing pfs server ?

But we can also decide to have one dedicated RasPI per sm. 

 

 Comments   
Comment by arnaud.lefur [ 05/Nov/18 ]

According to cloomis there is a vm server in the 5th rack

Comment by cloomis [ 03/Jan/19 ]

There will also be one fmsActor per SM. I had thought that with just the enuActor it would not be worth having a host/VM per SM, but with two actors it might be.

The point being that it would be nice to have the enu and fms actor instances self-identify themselves based on their hostnames.

The choice of whether to run VMs in the 5th rack or on the existing CB2F boxes, or to run the actors directly on dedicated SM computers in the IR[1-4] racks depends only on network/power topology. I believe that the only thing we really care about is being sure that the shutters do not stay open if Subaru networking or power are lost. I suppose we don't want any lamps/LEDs to stay on either. Thoughts, Yoshida, Hiroshige?

Comment by hassan [ 24/Jul/19 ]

Yoshida, Hiroshige: any feedback on this ticket?

Comment by hassan [ 25/Jul/19 ]

Craig will speak to Yoshida, Hiroshige during the MCS eng run to determine the best approach.

Comment by hassan [ 22/Nov/19 ]

Replaced by INSTRM-806. Close.

Generated at Sat Feb 10 16:26:19 JST 2024 using Jira 8.3.4#803005-sha1:1f96e09b3c60279a408a2ae47be3c745f571388b.