[PATCH pve-docs v2 1/1] ceph: add warning about mixing CRUSH rules with and without device classes when using autoscaler

[PATCH pve-docs v2 1/1] ceph: add warning about mixing CRUSH rules with and without device classes when using autoscaler

[Nicolas Frey]

Suggested-by: Friedrich Weber <f.weber@proxmox.com>
Signed-off-by: Nicolas Frey <n.frey@proxmox.com>
---
Changes since v1 (thanks Kefu!):
* Use Ceph's terminology for CRUSH rules with device classes
* adjust phrasing, as the autoscaler still scales on non-overlapping
    pools

 pveceph.adoc | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/pveceph.adoc b/pveceph.adoc
index 2aae6d6..b41bac9 100644
--- a/pveceph.adoc
+++ b/pveceph.adoc
@@ -916,6 +916,10 @@ TIP: If the pool already contains objects, these must be moved accordingly.
 Depending on your setup, this may introduce a big performance impact on your
 cluster. As an alternative, you can create a new pool and move disks separately.

+WARNING: When using the PG autoscaler, all pools in the cluster are required to
+use CRUSH rules of the same kind, either all specifying a device class,
+or none. Otherwise the autoscaler will skip the affected
+pools and their `pg_num` will not be adjusted.

 Ceph Client
 -----------
--
2.47.3

Re: [PATCH pve-docs v2 1/1] ceph: add warning about mixing CRUSH rules with and without device classes when using autoscaler

[Kefu Chai]

Nicolas Frey <n.frey@proxmox.com> writes:

> Suggested-by: Friedrich Weber <f.weber@proxmox.com>
> Signed-off-by: Nicolas Frey <n.frey@proxmox.com>
> ---
> Changes since v1 (thanks Kefu!):
> * Use Ceph's terminology for CRUSH rules with device classes
> * adjust phrasing, as the autoscaler still scales on non-overlapping
>     pools
>
>  pveceph.adoc | 4 ++++
>  1 file changed, 4 insertions(+)
>
> diff --git a/pveceph.adoc b/pveceph.adoc
> index 2aae6d6..b41bac9 100644
> --- a/pveceph.adoc
> +++ b/pveceph.adoc
> @@ -916,6 +916,10 @@ TIP: If the pool already contains objects, these must be moved accordingly.
>  Depending on your setup, this may introduce a big performance impact on your
>  cluster. As an alternative, you can create a new pool and move disks separately.
>
> +WARNING: When using the PG autoscaler, all pools in the cluster are required to
> +use CRUSH rules of the same kind, either all specifying a device class,
> +or none. Otherwise the autoscaler will skip the affected
> +pools and their `pg_num` will not be adjusted.
>
>  Ceph Client
>  -----------
> --
> 2.47.3