Hey folks, this week we discussed a few areas:
- sepia lab hardware
* some NICs and disks are dying more often in some older machines. decided to keep replacing these components since the rest of the boxes are still going strong.
* lab may be moved to a different datacenter next year, IPv6 testing would be possible with upgraded switches there
- ovh use
* We looked through areas where we still use OVH VMs for static services rather than bare metal, and determined they could almost all be moved out of the cloud, freeing up foundation budget - plan to do so over the next year
* Dan suggested looking into CDN services to help out download.ceph.com - will discuss more next week
- feedback mgr module
* this is a way for users to report bugs directly from the dashboard or cli
* demoed at the user + dev meeting on 12-16, looking for more feedback from users in the future
* uses a single redmine API key, so all bugs from a cluster are filed by the same user currently
* will be released with quincy
- scale testing in the sepia lab
* 1000 OSD cluster of gibba nodes
* dashboard and core team using it so far, other devs encouraged to get involved
* low resources per OSD to get to the desired scale, so not suitable for performance testing
* some issues due to low memory, cpu - trying some osd tuning to avoid these
* coordinate in #ceph-gibba irc channel on oftc
Josh
_______________________________________________ Dev mailing list -- dev@xxxxxxx To unsubscribe send an email to dev-leave@xxxxxxx
