Several file systems on a Security Analytics appliance will always show higher utilization that others. This is not necessarily a bad thing and shows that the appliance is working as designed.
On a Security Analytics capture appliance (not a Central Manager), two file systems will always be 100% full or close to it: "/pfs" and "/etc/solera/flows". These two file systems will normally match exactly.
[root@hostname ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/sda4 4.9G 3.0G 1.6G 66% /
/dev/sda6 2.0G 458M 1.4G 25% /gui
/dev/sda2 68G 5.0G 59G 8% /var
/dev/sda1 1.5G 41M 1.4G 3% /boot
/dev/sda5 4.9G 1.5G 3.2G 32% /ds
tmpfs 63G 228M 63G 1% /dev/shm
/dev/sda3 2.7T 3.2G 2.5T 1% /home
/dev/sdc1 21T 21T 0 100% /pfs
/dev/sdb1 9.6T 7.9T 1.6T 83% /var/lib/solera/meta1
/dev/sdb2 9.6T 7.9T 1.6T 83% /var/lib/solera/meta2
gaugefs 21T 21T 0 100% /etc/solera/flows
Both /pfs and /etc/solera/flows are virtual filesystems where all space is pre-allocated. They will always show as 100% (or close to 100%) full when mounted. As the data is captured, the /pfs file system uses FIFO (first in first out) for managing new traffic. The oldest data will always be overwritten first.
If /pfs is not listed in the output of "df" on a capture appliance, the system may not be licensed. As root, run "service solera status" to determine why /pfs is not mounted. If the capture filesystem is not running, that command should return "unlicensed". If it returns any other status, or if assistance with licensing is needed, contact technical support.
If /etc/solera/flows is not listed in the output of "df" and the appliance is not a Central Manager, the gaugefs service is stopped. Contact technical support for assistance.
The index file systems (/var/lib/solera/meta*) are managed by the indexing services. Once they reach 80% full, the indexer will, as needed, remove old index data in order to keep them less than or equal to 83% full. When these filesystems grow to >83%, the indexing services will remove enough old data to lower the utilization back to 80%.
If an index filesystem shows >83% full for sustained periods of time, contact technical support for assistance.