From 987e9058fde69e5f4073053525b996011941b213 Mon Sep 17 00:00:00 2001 From: Tim Burke Date: Wed, 30 Aug 2023 12:54:28 -0700 Subject: [PATCH] Tim Burke candidacy for Swift PTL (Caracal) Change-Id: Ibd9ab22e4c9ea8d2d8a5695d73da85d5eafa11d4 --- candidates/2024.1/Swift/tburke@nvidia.com | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) create mode 100644 candidates/2024.1/Swift/tburke@nvidia.com diff --git a/candidates/2024.1/Swift/tburke@nvidia.com b/candidates/2024.1/Swift/tburke@nvidia.com new file mode 100644 index 00000000..53ecd5b8 --- /dev/null +++ b/candidates/2024.1/Swift/tburke@nvidia.com @@ -0,0 +1,22 @@ +It would be my pleasure to continue serving as Swift PTL for the +Caracal cycle. + +I forsee us grappling with several challenges over the next few cycles: + +Successful clusters expand until they hit a wall, whether that's due +to budget or physical space constraints. Swift will need to adapt to +operate in more constrained environments. + +Total disk capacity is not the only measure of a cluster. Network +bandwidth (both aggregate across the cluster as well as single-stream) +and requests per second drive much of the client experience. Swift will +need to be able to ration those between tenants, similar to how it can +enforce capacity quotas. + +Operators are not the only ones interested in metrics and storage +insights. Users want to know what their individual access patterns +are like and how that has changed over time. + +I'm excited to see how much we can achieve in Caracal. + +Tim Burke