Revision a6321765

b/doc/design-partitioned.rst
163 163 
Currently it's possible to define an instance policy that limits the
164 164 
minimum and maximum value for CPU, memory, and disk usage (and spindles
165 165 
and any other resource, when implemented), independently from each other. We
166 
extend the policy by allowing it to specify more specifications, where
167 
each specification contains the limits (minimum, maximum, and standard)
168 
for all the resources. Each specification has a unique priority (an
169 
integer) associated to it, which is used by ``hspace`` (see below).
166 
extend the policy by allowing it to contain more occurrences of the
167 
specifications for both the limits for the instance resources. Each
168 
specification pair (minimum and maximum) has a unique priority
169 
associated to it (or in other words, specifications are ordered), which
170 
is used by ``hspace`` (see below). The standard specification doesn't
171 
change: there is one for the whole cluster.
170 172 

  		





  171
  173
  
    
For example, a policy could be set up to allow instances with this


  		





  172
  174
  
    
constraints:


  		





  ......




  183
  185
  
    
policy constraints, unless the flag ``--ignore-ipolicy`` is passed.


  		





  184
  186
  
    

  		





  185
  187
  
    
While the changes needed to check constraint violations are


  		





  186
  
  
    
straightforward, ``hspace`` behavior needs some adjustments. For both


  		





  187
  
  
    
standard and tiered allocation, ``hspace`` will start to allocate


  		





  188
  
  
    
instances using the specification with the highest priority, then it


  		





  189
  
  
    
will fall back to second highest priority, and so on. For tiered


  		





  190
  
  
    
allocation, it will try to lower the most constrained resources (without


  		





  191
  
  
    
breaking the policy) before going to the next specification.


  		





  
  188
  
    
straightforward, ``hspace`` behavior needs some adjustments for tiered


  		





  
  189
  
    
allocation. ``hspace`` will start to allocate instances using the


  		





  
  190
  
    
maximum specification with the highest priority, then it will try to


  		





  
  191
  
    
lower the most constrained resources (without breaking the policy)


  		





  
  192
  
    
before moving to the second highest priority, and so on.


  		





  192
  193
  
    

  		





  193
  194
  
    
For consistent results in capacity calculation, the specifications


  		





  194
  195
  
    
inside a policy should be ordered so that the biggest specifications


  		












Also available in:
  Unified diff