Data availability in a constrained deployment of a high-availability system in the presence of pending faults

A non-transitory computer-readable medium comprising instructions executable, by a processing resource of a high-availability (HA) storage system, to:

• receive an indication of a pending fault impacting a first failure domain of the HA storage system when the HA storage system is in a first configuration, the first configuration comprising a first arbiter node, of the first failure domain, to serve as an active arbiter to facilitate maintaining a quorum for the HA storage system, a first storage node of the first failure domain, and a second storage node of a second failure domain;
• wherein a second arbiter node of the second failure domain is to serve as a backup arbiter;
• responsive to the indication, initiate establishment of a second configuration for the HA storage system, the second configuration having a plurality of members including the first storage node, the second storage node, and the second arbiter node;
• and responsive to a majority of the plurality of members installing the second configuration, enable the second arbiter node to serve as the active arbiter by transferring state information to the second arbiter node.

A method comprising:

• receiving an indication of a pending fault impacting a first failure domain of a constrained deployment of a high-availability (HA) storage system serving data in a first configuration;
• wherein the HA storage system comprises a plurality of nodes each having a respective assigned weight value;
• wherein the HA storage system comprises a first failure domain comprising first and second nodes of the plurality of nodes, and a second failure domain comprising third and fourth nodes of the plurality of nodes;
• and
• wherein, in the first configuration, the respective assigned weight value of the second node of the first failure domain is greater than each respective assigned weight value of each of the other nodes of the plurality of nodes;
• and
• responsive to the indication, causing establishment of a second configuration for the HA storage system, in which the respective assigned weight value of the fourth node is greater than the respective assigned weight values of each of the other nodes of the plurality of nodes, by shifting a portion of the respective assigned weight value of the second node to the respective assigned weight value of the fourth node, wherein after impact of the pending fault, the HA storage system maintains an ability to serve data as a result of the third node and the fourth node having a quorum.