On Thu, Apr 18, 2019 at 3:47 PM Wesley Dillingham <[email protected]> wrote: > > I am trying to determine some sizing limitations for a potential iSCSI > deployment and wondering whats still the current lay of the land: > > Are the following still accurate as of the ceph-iscsi-3.0 implementation > assuming CentOS 7.6+ and the latest python-rtslib etc from shaman: > > Limit of 4 gateways per cluster (source: > https://access.redhat.com/documentation/en-us/red_hat_ceph_storage/3/html/block_device_guide/using_an_iscsi_gateway#requirements)
Yes -- at least that's what's tested. I don't know of any immediate code-level restrictions, however. You can technically isolate a cluster of iSCSI gateways by configuring them to access their configuration from different pools. Of course, then things like the Ceph Dashboard won't work correctly. > Limit of 256 LUNS per target (source: > https://github.com/ceph/ceph-iscsi-cli/issues/84#issuecomment-373359179 ) > there is mention of this being updated in this comment: > https://github.com/ceph/ceph-iscsi-cli/issues/84#issuecomment-373449362 per > an update to rtslib but I still see the limit as 256 here: > https://github.com/ceph/ceph-iscsi/blob/master/ceph_iscsi_config/lun.py#L984 > wondering if this is just an outdated limit or there is still valid reason to > limit the number of LUNs per target Still a limit although it could possibly be removed. Until recently, it was painfully slow to add hundreds of LUNs, so assuming that has been addressed, perhaps this limit could be removed -- it just makes testing harder. > Limit of 1 target per cluster: > https://github.com/ceph/ceph-iscsi-cli/issues/104#issuecomment-396224922 SUSE added support for multiple targets per cluster. > > Thanks in advance. > > > > > > Respectfully, > > Wes Dillingham > [email protected] > Site Reliability Engineer IV - Platform Storage / Ceph > > _______________________________________________ > ceph-users mailing list > [email protected] > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com -- Jason _______________________________________________ ceph-users mailing list [email protected] http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
