python-memcached get_multi has much faster than get when get multiple value

[Choe, Cheng-Dae]

nova use memcache with python.memcached's get function.

when multiple litem reterived it uses as for .. in .. loop..
in this case get_multi has better performance.

In my case, here is test result

get 2.3020670414
get_multi 0.0353858470917

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/211250

[Choe, Cheng-Dae (whitekid)]

If you have much more hypervisors, (we have 150 hypervisors), because if this issue Extended_availability_zone nova extensions has poor performance..

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/215974
Committed: https://git.openstack.org/cgit/openstack/oslo.cache/commit/?id=0a7c3bf46b2c33f5e469debc018cea79b025c632
Submitter: Jenkins
Branch: master

commit 0a7c3bf46b2c33f5e469debc018cea79b025c632
Author: Davanum Srinivas <email address hidden>
Date: Sat Aug 22 22:17:26 2015 -0400

    Implement get/set/delete multiple keys for dictionary backend

    For some Nova scenarios, we will need suppot for
    set_multi/get_multi/delete_multi for writing to either
    memcache or dictionary backend. So let's add support for
    operating on multiple keys. Note that if we don't implement
    this the default CacheBackend class throws NotImplementedError

    Closes-Bug: #1483322
    Change-Id: I1aca9e65b3be038a507ced8dd039c5413d6c4ac2

[OpenStack Infra (hudson-openstack)]

Change abandoned by Matt Riedemann (<email address hidden>) on branch: master
Review: https://review.openstack.org/211250
Reason: Looks like this isn't being worked on.

[Matt Riedemann (mriedem)]

Change https://review.openstack.org/#/c/203049/ is being used to integrate oslo.cache with nova which has the multi support since 0.7.0.

[OpenStack Infra (hudson-openstack)]

Change abandoned by Michael Still (<email address hidden>) on branch: master
Review: https://review.openstack.org/203049
Reason: This has been idle for ages. Please restore when you're ready to go again.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/274468

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/274468
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=205fb7c8b34e521bdc14b5c3698d1597753b27d4
Submitter: Jenkins
Branch: master

commit 205fb7c8b34e521bdc14b5c3698d1597753b27d4
Author: Davanum Srinivas <email address hidden>
Date: Fri Jan 29 12:50:58 2016 -0500

    Switch to oslo.cache lib

    Common memorycache was replaced by analogous tool
    from oslo.cache lib. In-memory cache was replaced
    by oslo.cache.dict backend. Memcached was replaced
    by dogpile.cache.memcached backend.

    Implements blueprint oslo-for-mitaka

    Closes-Bug: #1483322
    Co-Authored-By: Sergey Nikitin <email address hidden>
    Co-Authored-By: Pavel Kholkin <email address hidden>

    Change-Id: I371f7a68e6a6c1c4cd101f61b9ad96c15187a80e

[Doug Hellmann (doug-hellmann)]

This issue was fixed in the openstack/nova 13.0.0.0b3 development milestone.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/oslo.cache 0.7.0 release.