HPC infrastructure measures dataset - diskio table [2/13] part 1/2
收藏资源简介:
HPC infrastructure measures dataset - diskio table This record contains the part 1/2 of the diskio table of the HPC infrastructure measures dataset. Column descriptions: timestamp: The time at which the measurement was recorded. host: The hostname of the machine that generated the metric. name: The name of the block device, such as sda, sdb, or nvme0n1, for which the statistics were collected. weighted_io_time: The cumulative weighted time spent processing I/O requests, measured in milliseconds. merged_writes: The total number of write requests that were merged by the operating system or block layer before reaching the device. reads: The total number of completed read operations. writes: The total number of completed write operations. read_bytes: The total number of bytes read from the block device. write_bytes: The total number of bytes written to the block device. read_time: The cumulative time, in milliseconds, spent servicing read requests on the device. write_time: The cumulative time, in milliseconds, spent servicing write requests on the device. iops_in_progress: The number of I/O operations currently in progress at the time of collection. merged_reads: The total number of read requests that were merged before being sent to the device. io_time: The cumulative time, in milliseconds, during which the device had at least one I/O request in progress. io_util: The percentage of time during which the device was busy processing I/O operations. io_await: The average time, in milliseconds, that I/O requests took to complete. io_svctm: The average service time, in milliseconds, spent by the device processing each I/O request.



