On Mon, Apr 6, 2020 at 11:29 AM <ulrich.weigand@de.ibm.com> wrote:
Hi Sage,
I can provide some estimates and general guidance.
Thanks, this is really helpful!
First, the test suites are currently targetted to run on 'smithi' nodes, which are relatively low-powered x86 1u machines with a single NVMe divided into 4 scratch LVs (+ an HDD for boot + logs). (This is somewhat arbitrary--it's just the hardware we picked so the tests are written to target that.)
What's the size of those LVs? Do the tests require a particular minimum size?
see more about the hardware here https://wiki.sepia.ceph.com/doku.php?id=hardware:smithi cmeno@smithi110:~$ sudo pvs PV VG Fmt Attr PSize PFree /dev/nvme0n1 vg_nvme lvm2 a-- <372.61g 100.00m cmeno@smithi110:~$ sudo lvs LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert lv_1 vg_nvme -wi-a----- 89.40g lv_2 vg_nvme -wi-ao---- 89.40g lv_3 vg_nvme -wi-ao---- 89.40g lv_4 vg_nvme -wi-ao---- 89.40g lv_5 vg_nvme -wi-ao---- 14.90g cmeno@smithi110:~$ cheers, Christina
I've been telling the aarch64 folks that we probably want at least 25-50 similarly-sized nodes in order to run the test suites in a reasonable amount of time (e.g., minimal rados suite ~day and not days).
I'm not really sure how this maps on the Z hardware, but hopefully this provides some guidance!
Yes, it does! We'll certainly have to tune this is a bit depending on how everything performs on Z hardware, but this gives us a starting point.
Bye, Ulrich _______________________________________________ Dev mailing list -- dev@ceph.io To unsubscribe send an email to dev-leave@ceph.io