From 9168792e51fc3de01d86f94f5077a2e4fec5bd40 Mon Sep 17 00:00:00 2001 From: "Joshua M. Boniface" Date: Thu, 21 Sep 2023 22:39:00 -0400 Subject: [PATCH] Fix image link --- docs/architecture/georedundancy.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/architecture/georedundancy.md b/docs/architecture/georedundancy.md index df4794c..fe4777a 100644 --- a/docs/architecture/georedundancy.md +++ b/docs/architecture/georedundancy.md @@ -44,7 +44,7 @@ PVC clusters are quite network-intensive, as outlined in the [hardware requireme The storage write process for PVC is heavily dependent on network latency. -[![Ceph Write Process](images/ceph-write-process.png)](images/ceph-write-process.png) +[![Ceph Write Process](images/pvc-ceph-write-process.png)](images/pvc-ceph-write-process.png) As illustrated in this diagram, a write will only be accepted by the client once it has been successfully written to at least `min_copies` OSDs, as defined by the pool replication level (usually 2). Thus, the latency of network communications between the client and another node becomes a major factor in storage performance for writes, as the write cannot complete without at least 4x this latency (send, ack, recieve, ack). Significant distances and thus latencies (more than about 3ms) begin to introduce significant performance degredation, and latencies above about 5ms can result in a significant drop in performance.