Use case 4: Create threshold rules for specific filesets
This section describes the threshold use case to create threshold rules for specific fileset.
Consider that the initial cluster file system and fileset have the following
setting:
[root@gpfsgui-11 ~]# mmlsfileset nfs_shareFS
Filesets in file system 'nfs_shareFS':
Name Status Path
root Linked /gpfs/nfs_shareFS
nfs_shareFILESET Linked /gpfs/nfs_shareFS/nfs_shareFILESET
test_share Linked /gpfs/nfs_shareFS/test_share
Per default, the following rules are defined and enabled in the cluster:
[root@gpfsgui-11 ~]# mmhealth thresholds list
### Threshold Rules ###
rule_name metric error warn direction filterBy groupBy sensitivity
---------------------------------------------------------------------------------------------------------------
InodeCapUtil_Rule Fileset_inode 90 80 None gpfs_cluster_name,
gpfs_fs_name,
gpfs_fset_name 300
DataCapUtil_Rule DataPool_capUtil 90.0 80.0 high gpfs_cluster_name,
gpfs_fs_name,
gpfs_diskpool_name 300
MemFree_Rule mem_memfree 50000 100000 low node 300
MetaDataCapUtil_Rule MetaDataPool_capUtil 90.0 80.0 high gpfs_cluster_name,
gpfs_fs_name,
gpfs_diskpool_name 300
- Create new inode capacity usage rules for the specific filesets.
- To create a threshold rule for all filesets in an individual file system, use the following
command:
[root@gpfsgui-11 ~]# mmhealth thresholds add Fileset_inode --errorlevel 85.0 --warnlevel 75.0 --direction high --filterby 'gpfs_fs_name=nfs_shareFS' --groupby gpfs_cluster_name,gpfs_fs_name,gpfs_diskpool_name --sensitivity 300 --hysteresis 5.0 --name rule_ForAllFsets_inFS New rule 'rule_ForAllFsets_inFS' is created. The monitor process is activated
- To create a threshold rule for individual fileset, use the following
command:
[root@gpfsgui-11 ~]# mmhealth thresholds add Fileset_inode --errorlevel 70.0 --warnlevel 60.0 --direction high --filterby 'gpfs_fs_name=nfs_shareFS, gpfs_fset_name=nfs_shareFILESET' --groupby gpfs_cluster_name,gpfs_fs_name,gpfs_fset_name --sensitivity 300 --hysteresis 5.0 --name rule_SingleFset_inFS New rule 'rule_SingleFset_inFS' is created. The monitor process is activated
Note: In this case, for the nfs_shareFILESET fileset, you have specified both, the file system name and the fileset name in the filter.
The mmhealth thresholds add command gives an output similar to the following:# mmhealth thresholds list ### Threshold Rules ### rule_name metric error warn direction filterBy groupBy sensitivity ---------------------------------------------------------------------------------------------------------------------------------- InodeCapUtil_Rule Fileset_inode 90 80 None gpfs_cluster_name, 300 gpfs_fs_name, gpfs_fset_name DataCapUtil_Rule DataPool_capUtil 90.0 80.0 high gpfs_cluster_name, 300 gpfs_fs_name, gpfs_diskpool_name MemFree_Rule mem_memfree 50000 100000 low node 300 rule_SingleFset_inFS Fileset_inode 70.0 60.0 high gpfs_fs_name =nfs_shareFS, gpfs_cluster_name, 300 gpfs_fset_name =nfs_shareFILESET gpfs_fs_name, gpfs_fset_name rule_ForAllFsets_inFS Fileset_inode 85.0 75.0 high gpfs_fs_name =nfs_shareFS gpfs_cluster_name, 300 gpfs_fs_name, gpfs_fset_name MetaDataCapUtil_Rule MetaDataPool_capUtil 90.0 80.0 high gpfs_cluster_name, 300 gpfs_fs_name, gpfs_diskpool_name
- To create a threshold rule for all filesets in an individual file system, use the following
command:
- Run the mmhealth thresholds list command to list the individual rules'
priorities. In this example, the
rule_SingleFset_inFS
rule has the highest priority for thenfs_shareFILESET
fileset. Therule_ForAllFsets_inFS
rule has the highest priority for the other filesets that belong to thenfs_shareFS
file system, and theInodeCapUtil_Rule
rule is valid for all the remaining filesets.[root@gpfsgui-11 ~]# mmhealth thresholds list -v -Y|grep priority mmsysmonc_thresholdslist::0:1::InodeCapUtil_Rule:priority:4: mmsysmonc_thresholdslist::0:1::DataCapUtil_Rule:priority:4: mmsysmonc_thresholdslist::0:1::MemFree_Rule:priority:2: mmsysmonc_thresholdslist::0:1::rule_SingleFset_inFS:priority:1: mmsysmonc_thresholdslist::0:1::rule_ForAllFsets_inFS:priority:2: mmsysmonc_thresholdslist::0:1::MetaDataCapUtil_Rule:priority:4: [root@gpfsgui-11 ~]#