Switch over to secondary Cortex

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • Gumby
    Moderator
    • May 2004
    • 437

    Switch over to secondary Cortex

    I have a secondary Cortex running on a Win 7 Pro VM on a Mac Mini. It is just about OK for emergencies but suffers from increasing button lag with duration of control. I blame this on the VM. Soon it will have a proper home - in general it is very useful to have a secondary, eg castastrophic hardware failures.

    However, there are 2 issues that may not be related to the VM:

    1) You cannot stop the paused network on the secondary without making it run first. This is a nuisance, because...

    2) On switchover there is a period of disco inferno when all the lights flash on and off. While it may be useful to know that the secondary has taken control, it doesn't go down well in the middle of the night.

    3) End state with secondary active can leave "Manual control only" lights in opposite to their original state, eg on when they were off (in the middle of the night).

    I may have an unusual failure mode on the primary, the desktop is unresponsive, I can ping but not RDC in. I think the primary is no longer servicing the network, based on comms window on the secondary and also network watchdog firing into reflex mode if secondary is disabled.

    Anyone have similar? WAF currently negative :-(
    ----------------------
    www.gumbrell.com
  • cliffwright
    Automated Home Guru
    • Mar 2007
    • 117

    #2
    Very old thread here I know (Most are these days - why so quiet?) ... but ...

    I've just got a secondary machine that lies largely dormant (Ironically this one failed recently with the HDD dying on me so I've just had to re-build it) ... with the idea being that if the Primary failed, I'd physically go and swap over by swapping the USB cable into the USB PCU from the Primary to the secondary machine. I'm happy with level of resilience because my primary has proven so reliable and can suck up the 5-10 mins downtime that'd cause.

    I use a Synology NAS to take regular backups of the Primary Cortex files and sync these across to the secondary machine so any swap over should be seamless functionality wise.

    Have you progressed any at all since this original thread? Interested to hear what setup you've got ... are you saying you've got both machines connected to the Idranet, with 1 "Paused" and the other active ... and you switch over is to swap the network states to toggle from 1 machine to the other?
    www.clifford-wright.co.uk/blog

    Comment

    • Karam
      Automated Home Legend
      • Mar 2005
      • 863

      #3
      The first post is describing a situation where two Cortex platforms are connected to the same IDRANet with both being in an active state, though one is 'paused'. This allows automatic handover - such that the network can continue to function without user intervention. There are a number of reasons why it may not be a seamless transfer. One is because you don't necessarily have to use the same database on the two platforms, for example the secondary might have less licenced options. Another reason is that aside from the physical states of the devices out there in the world, Cortex has an internal state representing many different variables related to these or other abstracted internal logic. For example the current count value of timers, the reasons why a light or other appliance might presently be in a particular state (e.g current override mode), and so on. You might describe this at its 'mental view'. To get a totally seamless transfer you'd need to transfer (or maintain) this internal mental state to the secondary platform and this can't be deduced by the secondary platform simply monitoring network traffic. So in a nutshell it is a feature which allows continuation of functionality without user intervention (e.g user might be away from home or not someone who knows how to perform a manual switch over), but imperfect in terms of state continuity after the handover process. Of course it should be remembered that this is a backup mechanism and not something which should be called upon regularly. If it is, then it points to other issues which require resolution.

      Comment

      Working...
      X