Recovering the hot clusters and the deletion log are quick; it takes about 30 seconds to read 2500 hot clusters and a log of 3000 deletion entries into memory. Thus, on a system crash with a 18 GB disk, Hummingbird can start to service requests in 30 seconds, while UFS will take more than 20 minutes to perform the necessary fsck before requests can be serviced. While rebuilding the in-memory meta-data (Step 6 in Section 3.4), the FS read time increased a small amount (e.g., from 2.03 ms to 2.50 ms for a 8 GB disk).
Journalling file systems will recover integrity quickly compared with traditional UFSs due to log-based recovery. However, journalling alone will not fetch hot data from disk as part of the recovery process.