drive read latency is greater than 100s when leader handle fetch data request #760
Description
This happened when testing 16M blob r/w, details in issue60
Leader (sm-long-running2-3) metrics:
HS side: https://console.sherlock.io/d/b43364d3-4b8c-476b-a7e8-73219b91f82e/nuobject-storage-manager?orgId=1&var-namespace=sdsapp&var-subcluster=nuobject2-sh-3&var-pg=$__all&var-sm=$__all&from=2025-06-13T12:10:00.000Z&to=2025-06-13T12:50:39.000Z&timezone=browser
Node side (disk is sdd):
https://console.sherlock.io/d/nLt4mwcVz/node-status-dashboard-tess?orgId=1&from=2025-06-13T12:10:00.000Z&to=2025-06-13T12:50:39.000Z&timezone=browser&var-node=tess-node-ltl7f-tess30.stratus.lvs.ebay.com&var-Filters=