Under certain scenarios, the disk attempts to read past the end. See
|
// eprintln!( |
|
// "[append] nvme::ops::write failed: zone={}, index={}, addr={}, size={} bytes. Error: {:?}", |
|
// chunk_location.zone, |
|
// chunk_location.index, |
|
// write_addr, |
|
// mut_data.len(), |
|
// err |
|
// ); |
|
// println!("[append] writing chunk to {} bytes at addr {}", chunk_location.zone, write_addr); |
This seems to happen when we make enough queries that the disk fills up.
Probably related to eviction not being finished at the moment
Under certain scenarios, the disk attempts to read past the end. See
OxCache/oxcache/src/device.rs
Lines 435 to 442 in 78cc71f
OxCache/oxcache/src/device.rs
Line 415 in 78cc71f
This seems to happen when we make enough queries that the disk fills up.
Probably related to eviction not being finished at the moment