nova/nova/scheduler
Sahid Orentino Ferdjaoui f254729a56 scheduler: fix _get_sharing_providers to support unlimited aggr
A Request-Too-Long error can be raised in case of too many aggregates
are attached to a host. The fix is to split the requests.

The request is executed by host via a periodic task. Result is set in
cache so no performances impact is expected.

The change introduce a new compute configuration option
`compute.sharing_providers_max_uuids_per_request = 200`.

Closes-bug: #2068893

Change-Id: I1355c0813cb99067ead60f9835938566289beb20
Signed-off-by: Sahid Orentino Ferdjaoui <sahid.ferdjaoui@industrialdiscipline.com>
2024-07-03 12:58:04 +00:00
..
client scheduler: fix _get_sharing_providers to support unlimited aggr 2024-07-03 12:58:04 +00:00
filters scheduler: AggregateMultitenancyIsolation to support unlimited tenant 2024-04-26 07:07:03 +00:00
weights [codespell] fix final typos and enable ci 2023-12-15 12:32:42 +00:00
__init__.py Improve hacking rule to avoid author markers 2014-05-05 14:35:20 +02:00
host_manager.py Filter PCI pools based on Placement allocation 2022-10-17 13:56:18 +02:00
manager.py Make scheduler lazy-load the placement client 2023-03-22 08:59:25 -07:00
request_filter.py Packed virtqueue support was added. 2023-11-29 16:06:33 +02:00
rpcapi.py Remove return from rpc cast 2022-06-18 16:23:26 +05:30
utils.py libvirt: Support maxphysaddr. 2024-02-27 14:16:25 +09:00