Description of problem: During a continuous write workload, writes were accelerated by the cache layer up until the low watermark was reached. At this point performance to the client degraded to pre-cache levels. Version-Release number of selected component (if applicable): How reproducible: Tried several times, and each time the effect was observed Steps to Reproduce: 1. create a tier volume with SSD as the tier 2. create files on the volume with dd repeatedly 3. observe throughput degradation as the tier > 60% utilised Actual results: client writes degrade considerably Expected results: client writes should degrade, but I saw 150MB/s become 4MB/s Additional info: The test environment I used was vm's under KVM. Each gluster node has an SSD disk for hot, and a replica 2 for cold tier.
This bug is getting closed because GlusteFS-3.7 has reached its end-of-life. Note: This bug is being closed using a script. No verification has been performed to check if it still exists on newer releases of GlusterFS. If this bug still exists in newer GlusterFS releases, please reopen this bug against the newer release.