_lm_asm_enq_hashing _ges_diagnostics_asm_dump_level _asm_runtime_capability_volume_support _asm_disable_multiple_instance_check _asm_disable_amdu_dump _asmsid _asm_allow_system_alias_rename _asm_instlock_quota asm_diskstring _asm_disk_repair_time asm_preferred_read_failure_groups _asm_disable_profilediscovery _asm_imbalance_tolerance _asm_shadow_cycle _asm_primary_load_cycles _asm_primary_load _asm_secondary_load_cycles _asm_secondary_load asm_diskgroups asm_power_limit _asm_log_scale_rebalance _asm_sync_rebalance _asm_ausize _asm_blksize _asm_acd_chunks _asm_partner_target_disk_part _asm_partner_target_fg_rel _asm_automatic_rezone _asm_rebalance_plan_size _asm_rebalance_space_errors _asm_libraries _asm_maxio _asm_allow_only_raw_disks _asm_fob_tac_frequency _asm_emulate_nfs_disk _asmlib_test _asm_allow_lvm_resilvering _asm_lsod_bucket_size _asm_iostat_latch_count _asm_disable_smr_creation _asm_wait_time _asm_skip_resize_check _asm_skip_rename_check _asm_direct_con_expire_time _asm_check_for_misbehaving_cf_clients _asm_diag_dead_clients _asm_reserve_slaves _asm_kill_unresponsive_clients _asm_disable_async_msgs _asm_stripewidth _asm_stripesize _asm_random_zone _asm_serialize_volume_rebalance _asm_force_quiesce _asm_dba_threshold _asm_dba_batch _asm_usd_batch _asm_fail_random_rx _asm_max_redo_buffer_size _asm_max_cod_strides _asm_kfioevent _asm_evenread _asm_evenread_alpha _asm_evenread_alpha2 _asm_evenread_faststart _asm_dbmsdg_nohdrchk _asm_root_directory _asm_hbeatiowait _asm_hbeatwaitquantum _asm_repairquantum _asm_emulmax _asm_emultimeout _asm_kfdpevent _asm_storagemaysplit _asm_avoid_pst_scans _asm_compatibility _asm_admin_with_sysdba _asm_allow_appliance_dropdisk_noforce _asm_appliance_config_file _disable_rebalance_space_check _disable_rebalance_compact       _disable_rebalance_compact : Setting _DISABLE_REBALANCE_COMPACT=TRUE will disable the compacting phase of the disk group rebalance.   Since version 11.1, the ASM performs a disk compacting once the rebalance is complete. This may have a noticeable impact on the first rebalance after the upgrade from 10g. The idea of the compacting phase it to move the data as close the outer tracks of the disks (the lower numbered offsets) as possible. The first time the rebalance runs in 11g, it could take a while if the disk group configuration changed – for example after an ALTER DISKGROUP … ADD DISK. Subsequent manual rebalance without a configuration change should not take as much time. A disk group where the compacting phase of the rebalance has done a lot of work will tend to have better performance than the pre-compact disk group. The data should be clustered near the higher performing tracks of the disk, and the seek times should be shorter. Relevant initialization parameter: _DISABLE_REBALANCE_COMPACT. Relevant disk group attribute: _REBALANCE_COMPACT. The holemarks of the compacting phase:
  • Rebalance is taking ‘too long’
  • Updates to gv$asm_operation have stopped
  • In the ARB0 trace file we see many lines like this: ARB0 relocating file +<diskgroup>.nnn.mmm (1 entries)
  • Stack (systemstate, processstate or pstack) shows kfdCompact() function.
  _asm_avoid_pst_scans: If set, limit the number of full scans (of all disks) we do when reading the PST in kfdp_readMeta(). This is also required to activate split PST checks (which rely on the lock value block being available)     _asm_evenread - enable/disable even read. 0 = Defer to the disk group attribute (to be implemented) 1 = ENABLE (always on) 2 = DISABLE (always off) 18 = Enabled iff offline disks are found _asm_evenread_tracing - tracing; 0 is off, 1 is 1 line per IO, 3 is verbose _asm_evenread_alpha -alpha value, x2^14 (16 is a good value) _asm_evenread_alpha2 - secondary alpha value, x2^14 _asm_evenread_faststart - number of IOs after offline/online to use alpha2 (default 0)