OSD磁盘日常监控

摘要：对ceph OSD磁盘，做好定期的性能数据采集和通电时长管理，长期的数据积累对磁盘的性能与生命周期管理会有一定帮助,同时也能确保整个集群性能的稳定。

磁盘碎片管理

查看磁盘碎片

# xfs_db -c frag -r /dev/sdb1
actual 981, ideal 964, fragmentation factor 1.73%

整理碎片

# xfs_fsr /dev/sdb1

OSD磁盘性能

root@demo:~# ceph --admin-daemon /var/run/ceph/ceph-osd.0.asok perf dump
{
    "WBThrottle": {
        "bytes_dirtied": 0,
        "bytes_wb": 0,
        "ios_dirtied": 0,
        "ios_wb": 0,
        "inodes_dirtied": 0,
        "inodes_wb": 0
    },
    "filestore": {
        "journal_queue_max_ops": 300,
        "journal_queue_ops": 0,
        "journal_ops": 3210,
        "journal_queue_max_bytes": 33554432,
        "journal_queue_bytes": 0,
        "journal_bytes": 5007708,
        "journal_latency": {
            "avgcount": 3210,
            "sum": 9.786657669
        },
        "journal_wr": 2643,
        "journal_wr_bytes": {
            "avgcount": 2643,
            "sum": 13221888
        },
        "journal_full": 0,
        "committing": 0,
        "commitcycle": 270,
        "commitcycle_interval": {
            "avgcount": 270,
            "sum": 1352.249982669
        },
        "commitcycle_latency": {
            "avgcount": 270,
            "sum": 2.088010691
        },
        "op_queue_max_ops": 50,
        "op_queue_ops": 0,
        "ops": 3210,
        "op_queue_max_bytes": 104857600,
        "op_queue_bytes": 0,
        "bytes": 4962768,
        "apply_latency": {
            "avgcount": 3210,
            "sum": 34.241022045
        },
        "queue_transaction_latency_avg": {
            "avgcount": 3210,
            "sum": 0.431766161
        }
    },
......
}

root@demo:~# ceph osd perf
osd fs_commit_latency(ms) fs_apply_latency(ms)
  0                     3                   10
  1                     3                   12
  2                     3                   13

通电时长

# smartctl -A /dev/sda
smartctl 5.41 2011-06-09 r3365 [x86_64-linux-3.8.0-44-generic] (local build)
Copyright (C) 2002-11 by Bruce Allen, http://smartmontools.sourceforge.net


Current Drive Temperature:     27 C
Drive Trip Temperature:        65 C
Elements in grown defect list: 0
Vendor (Seagate) cache information
  Blocks sent to initiator = 0
Vendor (Seagate/Hitachi) factory information
  number of hours powered up = 53413.70
  number of minutes until next internal SMART test = 7