Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

1422

August 12th, 2015 23:00

FAST VP health check

Guys - Checking if someone assist me on monitoring pool when we implement FAST VP on vmax 40k frames

well, i have three pools(SSD, FC & SATA) and i always map dev/lun to FC pool....

but i always get alarms that SSD pool is 98% used. at this moment im unsure on how to react if this values is normal.

of course i understand that FAST-VP promotes and demotes LUN level data across storage tiers(EFD/FC/SATA tiers) to accomplish performance but I’m still checking below  SDD pool usage(94%) is still optimal.

                     S Y M M E T R I X   P O O L S                      

---------------------------------------------------------------------------

Pool         Flags Dev              Usable       Free       Used Full Comp

Name         PTECSL Config           Tracks     Tracks Tracks  (%)  (%)

------------ ------ ------------ ---------- ---------- ---------- ---- ----

DEFAULT_POOL SSF-EI RAID-6(6+2)   909480720   185803141   723677579   79    0

123_tp_SATA_1    TSFDEI RAID-6(6+2)  1133247552 585256860   547990692   48    0

123_tp_FC_1        TFFDEI RAID-6(6+2)   743429220   418883160   324546060   43    0

123_tp_SSD_1      TEFDEI RAID-6(6+2)   214999260   2193864      212805396    98    0

26 Posts

August 13th, 2015 05:00

If you have no TDEV bound to the flash, pool, then there is no cause for concern and I would change or disable your alerting in Unisphere on the SSD pool.  In general, under FASTVP best practices, you want the SSD pool full with all devices bound to FC pool.  It's a good thing!

82 Posts

August 14th, 2015 01:00

Thanks for the clarification but what will happen internally when SSD is reaches 100% ?

when do i justify if i need to add few new SSD's to the pool...

2) well, in below case(other frame) i see SSD usage is 87% and SATA is at 96% but no TDEV's are mapped to neither to SSD nor to SATA.

Pool Name Total (GB) Free (GB) Used (GB) Used %
DEFAULT_POOL SSF-EI RAID-6(6+2)   53859.6 40009.3 13850.4 25
RDFA_DSE_21  RFF-EI 2-Way Mir      1464.4 1464.4 0 0
321_tp_26_1 TFFDEI RAID-6(6+2)- FC  116869.4 89803.1 27066 23
321_tp_56_1 TEFDEI RAID-6(6+2)-SSD   2614 315.9 2298.1 87
321_tp_22_1 TSFDEI RAID-6(6+2)-SATA  101797.4 4053.8 97743.9 96

August 19th, 2015 14:00

If you are not binding anything to SSD drives and the pool  reserve capacity is set to 1% (I guess it is set to 2% in your case) the pool will never go to 100%.

This is normal behavior. Same is true  for SATA disks.

82 Posts

August 19th, 2015 22:00

Thanks for the info. bwt, how do i know if my pool reserve capacity is set to 1% or 2%? - is there any cmd to know.

August 20th, 2015 07:00

There are two different parameters where this is controlled.

First  is at Fast level and that is the global for all the thin pool.

It is supposed to be 10% and do not change this. To see what it is set to run this command

symfast -sid XXXX list  -control_parms |grep "Thin Pool Reserved Capacity(%)"

Make sure it is set to 10%

Second the parameter is based on each pool and here is where you want to set it to 1%

Command to see it is

symcfg -sid XXXX  show -pool -detail -thin |grep "Pool Reserved Capacity"

You can change this with symconfigure command

set pool , type thin, pool_resv_cap=

With unisphere the first capacity can be changed/looked at in fast VP setting

fastsetting.PNG.png

The individual setting can be changed  at storage>Thinpools >Pool_Name

Thinpool.PNG.png

Hope this answers your question

82 Posts

August 21st, 2015 04:00

Thanks Friends(Haralson & Rednirb).

No Events found!

Top