Backlog #2340

filter VM listing server-side

Added by Felix Schüren about 6 years ago. Updated almost 6 years ago.

Status:PendingStart date:09/26/2013
Priority:LowDue date:
Assignee:-% Done:

0%

Category:Core & System
Target version:-

Description

I would like the option to filter the API output server-side - often, I'm finding myself in need of just a few fields of all the data stored about a VM, and this is very inefficient at large numbers of VMs, if basically all I want is a list of just two or three columns of data.

To keep backwards compatibility, I would recommend to add an option to the listing queries to specify a list of columns I want to receive. Either use a single-value with a field separator, or use multiple fields, e.g.

a)
<includeColumns>vmId|assignedIp</includeColumns>

b)
<includeColumn>vmId</includeColumn>
<includeColumn>assignedIp</includeColumn>
[...]

this will make scaling a lot simpler for large installations.


Related issues

Related to Feature #2371: Paginate the .info API responses Closed 10/10/2013
Related to Backlog #2338: make VM listing scale better Pending 09/26/2013

History

#1 Updated by Ruben S. Montero about 6 years ago

  • Category set to Core & System

#2 Updated by Ruben S. Montero about 6 years ago

  • Target version changed from Release 4.4 to Release 4.6

#3 Updated by Ruben S. Montero almost 6 years ago

This will probably need to be implemented client-side. There are no processing problems currently in the core to get all the information, so there is no need to move client processing (filtering) to the core.

There are libraries and XML parsers that let you process the XML in chunks.

This is a problem of the current client API, so this performance improvement will be exposed there.

#4 Updated by Ruben S. Montero almost 6 years ago

  • Status changed from Pending to New

#5 Updated by Ruben S. Montero almost 6 years ago

  • Tracker changed from Feature to Backlog
  • Priority changed from Normal to Low
  • Target version deleted (Release 4.6)

#6 Updated by Ruben S. Montero almost 6 years ago

  • Status changed from New to Pending

#7 Updated by Ruben S. Montero almost 6 years ago

After some benchmarking, XML transport in the xml-rpc part is the most time consuming task. So pagination (server-side) will be implemented as a first step to filter large pools.

#8 Updated by Ruben S. Montero almost 6 years ago

  • Related to Feature #2371: Paginate the .info API responses added

#9 Updated by Ruben S. Montero almost 6 years ago

Also available in: Atom PDF