That sounds good! So do I get you right about the r/w access thing: default behavior will be "shared", and if I want to use one of the others I have to change it in each session? What I imagined was that it'd be possible to set a default behavior for all the console sessions.
Mit freundlichen Grüßen / Kind regards
Jonathan (Nathan) Hermann
IT Specialist
Global Technology Services / Data Center Services
Mobile: +49-160-98976942 IBM
Allee 1
E-mail: [email protected] Ehningen,
71139
Germany
IBM Deutschland Infrastructure Technology Services GmbH
Geschäftsführung: Ulrike Hetzel
Sitz der Gesellschaft: Ehningen
Registergericht: Amtsgericht Stuttgart, HRB 727973
From: Jarrod B Johnson <[email protected]>
To: xCAT Users Mailing list <[email protected]>
Date: 07.04.2014 16:08
Subject: Re: [xcat-user] console escape sequences to support
Ok, would be 'free' to do ctrl-e, c, p anyway (since the behavior already
is implemented as a default behavior on connect)
Do you think exclusive write would be an argument to client, or a sequence
to request? I suppose I could make ctrl-e, c, a "request exclusive-write"
access. I don't know if a sequence is needed to ask to 'allow shared
write' or just assume client wants either exclusive-write or read-only
until they disconnect... Also, if a sequence or argument is called for to
prevent others from reading your console as you type or not....
(Embedded image moved to file: pic20140.gif)Inactive hide details for
Jonathan Hermann ---04/07/2014 10:02:49 AM---Hi Jarrod, I'm using ctrl-e,
c, p a lot, would be reallyJonathan Hermann ---04/07/2014 10:02:49 AM---Hi
Jarrod, I'm using ctrl-e, c, p a lot, would be really glad if this could be
kept as
From: Jonathan Hermann <[email protected]>
To: xCAT Users Mailing list <[email protected]>
Date: 04/07/2014 10:02 AM
Subject: Re: [xcat-user] console escape sequences to support
Hi Jarrod,
I'm using ctrl-e, c, p a lot, would be really glad if this could be kept as
an explicit command. At times it's very helpful for debugging. And
especially for stateless nodes I find it way easier than to manually grep
the log files on the xCAT node.
Concerning r/w behaviour: In my experience (though limited, I admit) it
does make sense to have an exclusive-write option. Thinking of larger sites
where multiple persons are maintaining a cluster at the same time.
Mit freundlichen Grüßen / Kind regards
Jonathan (Nathan) Hermann
IT Specialist
Global Technology Services / Data Center Services
Mobile: +49-160-98976942 IBM
Allee 1
E-mail: [email protected]
Ehningen, 71139
Germany
IBM Deutschland Infrastructure Technology Services GmbH
Geschäftsführung: Ulrike Hetzel
Sitz der Gesellschaft: Ehningen
Registergericht: Amtsgericht Stuttgart, HRB 727973
From: Jarrod B Johnson <[email protected]>
To: "xCAT Users Mailing list" <[email protected]>
Date: 07.04.2014 15:52
Subject: [xcat-user] console escape sequences to support
So in the new console server, thus far all the client currently supports
is:
ctrl-e, c, ? (help)
ctrl-e, c, . (disconnect console)
I'm going to be adding:
ctrl-e, c, o (reconnect console, note that it already automatically
detects and corrects most situations that I employed ctrl-e, c, o to fixup
today, and that it accelerates fixup of an unconnected session
automatically on client connect)
ctrl-e, c, l, 0 (send break)
Already today, it automatically does something analagous to ctrl-e, c, p on
connection without user request, though I could add it as an explicit
command.
FWIW, the default behavior is currently all clients get write access. I
recall at least one person requesting the current behavior, but was
wondering how many of the following server-enforced modes would be desired:
shared (many can read and write)
exclusive-write (many can see, only one can write at a time)
exclusive-full (only one can read or write to a console at a given time,
buffered replay would be cleared on close of a client session)
Given that, what would be good escape sequences to represent requesting an
write-exclusive or full-exclusive session?
More generally, any other sequences that are very popular or else would
people prefer a scheme different from the conserver approach? I personally
have difficultly with the ubiquitous '~' scheme and find it somewhat handy
to not have to think how many ssh sessions I might have nested.
------------------------------------------------------------------------------
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test & Deployment
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees_APR
_______________________________________________
xCAT-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/xcat-user
------------------------------------------------------------------------------
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test & Deployment
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees_APR
_______________________________________________
xCAT-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/xcat-user
------------------------------------------------------------------------------
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test & Deployment
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees_APR
_______________________________________________
xCAT-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/xcat-user
<<attachment: pic20140.gif>>
------------------------------------------------------------------------------ Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test & Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees
_______________________________________________ xCAT-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/xcat-user
