Discussions
Categories
- 17.9K All Categories
- 3.4K Industry Applications
- 3.3K Intelligent Advisor
- 62 Insurance
- 536.1K On-Premises Infrastructure
- 138.2K Analytics Software
- 38.6K Application Development Software
- 5.7K Cloud Platform
- 109.4K Database Software
- 17.5K Enterprise Manager
- 8.8K Hardware
- 71.1K Infrastructure Software
- 105.2K Integration
- 41.6K Security Software
Size and number of luns for the database (500GB or 1TB)
Hi cracks,
We would like to move our production database to ASM (8,5TB)...
what is more preferable, luns of 1 TB (16) or luns of 500GB (32)?
thanks
Best Answer
-
what I would like to say is that is faster add a "small disk" (in a crisis time) because the data to rebalance is less, let me explain with an image:
Ok, I know that if you need to add another 512GB you will take the double of time, but I only think in the rebalance time operation and the affectation to the production database.
Answers
-
Preferable based on what criteria?
I suggest you review https://docs.oracle.com/cd/E11882_01/server.112/e18951/asmprepare.htm#OSTMG11000
To avoid kernel contention, a minimum of four LUNs (Oracle ASM disks) of equal size and performance is recommended for each disk group. Do you plan to have more than 4 disk groups? 1 TB HDD are usually faster than older 500 GB.
-
Hi @Dude!
Thanks for response, my plan is to have these 4 diskgroups:
But I would like to know what is preferable,
for example (assuming that the performance of the 500GB and 1TB is the same)
If I will need to increase the size of the database adding new disks, the adventaje of have 500GB is only that the re-balancing operation is faster?
-
Where do you store FRA?
Why do you think re-balance using 500 GB disks is faster? Based on what idea?
-
I will store FRA in RECO area (arch and the last full backup).
I will change the number of luns ;P
When you need to add a new disk the size of it is not relevant for the time of re-balance?
-
Your RECO changed from 4 x 500 GB to 24 x 1 TB?
As a rule of thumb FRA should at least be 3 times the size of your database.
When you need to add a new disk the size of it is not relevant for the time of re-balance?No, because ASM does not mirror disks, but files. How much data needs to be re-balanced when you add new disks depends on how much data is used, not the size of disks. The re-balance operation makes sure that your data is split among all disks evenly, including data mirroring requirements.
-
Thanks @Dude!
I will follow the rule of thumb and resize FRA
If I have 5,3TB of data used, please, can you clarify me the explanation that is faster 12x1TB of luns more than 24x500MB? Like sesame street because I need to explain to my bosses ;P
SQL> select tbs_size SizeGb, a.free_space FreeGb, tbs_size-a.free_space UsedGb
from (select round(sum(bytes)/1024/1024/1024 ,2) as free_space
from dba_free_space) a,
(select sum(bytes)/1024/1024/1024 as tbs_size
from dba_data_files) b; 2 3 4 5
SIZEGB FREEGB USEDGB
---------- ---------- -------------------------
8601.34863 3229.92 5371.42
-
Imagine you have 24 buckets of water and add one more. Then you take a cup of water out of each bucket to fill the empty one until each 25 buckets are filled evenly. Now try to imagine the same situation with 12 buckets or 1000 buckets. Which one is faster?
More devices do not necessarily translate to better performance because it can add to overhead or indirect computation time. ASM documentation recommends a minimum of 4 devices. Adding more from what I understand does not necessarily give better performance, otherwise the minimum would be higher.
However, the more devices you have, the more likely you increase the risk of hardware failure. Also keep in mind that you should only add devices of the same size and performance. If you need to substantially increase your storage capacity, it will be more economic the less devices you need to add.
-
In addition to all the other comments, keep in mind that a LUN is just an abstract layer on top what whatever physical disks may be in use. If you start with a SAN cabinet full of spinning disks, different SA's will configure a LUN differently. Some SA's will create a LUN by grouping specific physical disks, used solely by that LUN. Other SA's will group a bunch of disks into a storage pool, then allocate multiple LUNs from that pool. In that case, you'd have multiple LUNs all equally sharing the same physical disks. Other SAs will extend that concept an put the entire SAN cabinet into a single storage pool, allocating LUNs as needed.
-
we have the last option:Entire SAN cabinet with full SSD disks into a single storage pool.
Do o you recommend then create this configuration?
4 disks of 3TB for data diskgroup and
4 disks of 9TB for fra diskgroup
-
I don't see anything wrong with it.
