Blog: Ceph standalone and upstream
Blog articles about Ceph and use cases. A great technology by the community for use in production.
Ceph collocation and failure domains practice – advanced extravaganza
Concluding the series of collocation, this article provides additional special thoughts and considerations for limited and large Ceph cluster designs.
Collocation example with 2 sub-level failure domains and 3 top-level failure domains
What if failure domains aren’t that easy and have developed over time into a complex dependency ? The article gives an example illustrating such a “grown” infrastructure.
Collocation example – A recommended example with Erasure Coding 4+2 and 7 failure domains
Saving on capacity footprint – using Erasure Coding across multiple failure domains. Getting the right approach for redundancy and durability of data.
Collocation example – A simple example with 4 failure domains and replication of 3
Distribution of data and roles with Ceph across 4 failure domains – a collocation example.
Collocation example – A complex example with 2 failure domains and replication of 4
A simply common approach with 2 main DCs: an example for the calculation of the number of nodes for collocation in Ceph with a common pattern of nearby DCs inherited from legacy data center redundancy.
Collocation example – A complex example with 3 failure domains and replication of 3
The next example of calculation of the number of nodes for collocation in Ceph with respect to failure domains: a more complex cluster with 3 failure domains with replica 3
Collocation examples – Simple one: 3 failure domains – replication of 3
First article with examples of calculation of the number of nodes for collocation in Ceph with respect to failure domains: 3 failure domains with replica 3
Advanced Collocation – Placement of the additional Roles
As the last article about the foundation of calculation of the number of nodes for collocation, now the additional roles to place across failure domains are discussed.
Advanced Collocation – Data Placement Constraints
Taking the discussion about collocation and number of nodes in Ceph clusters forward, now with respecting failure domains the article discusses the data placement constraints.
Collocation and number of nodes in a Red Hat Ceph cluster
[ 23 min read ] Since a couple of versions, the Red Hat Ceph Storage product supports the collocation of instances of different roles on the same nodes. The instances for those roles cannot be placed randomly together but should follow rules. Those have been described in the docs and in a knowledge base article…