- Community Home
- >
- Storage
- >
- Entry Storage Systems
- >
- Disk Enclosures
- >
- 12H performance
Categories
Company
Local Language
Forums
Discussions
Forums
- Data Protection and Retention
- Entry Storage Systems
- Legacy
- Midrange and Enterprise Storage
- Storage Networking
- HPE Nimble Storage
Discussions
Discussions
Discussions
Forums
Forums
Discussions
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
- BladeSystem Infrastructure and Application Solutions
- Appliance Servers
- Alpha Servers
- BackOffice Products
- Internet Products
- HPE 9000 and HPE e3000 Servers
- Networking
- Netservers
- Secure OS Software for Linux
- Server Management (Insight Manager 7)
- Windows Server 2003
- Operating System - Tru64 Unix
- ProLiant Deployment and Provisioning
- Linux-Based Community / Regional
- Microsoft System Center Integration
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Community
Resources
Forums
Blogs
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО11-25-2001 03:05 AM
тАО11-25-2001 03:05 AM
I have a performance pb on the 12H i use the arrydsp -r for recommandation it show a recommandation to add a disk after doing this the recomondation is showing "no recommandation performance is optimal"but still pb in perf i use
arraydsp -m to get performance metrics the SCSI Q is very hihg (98,100 ...)
have you an idea to solve this problem.
LUN are used for raw device of Oracle database.
Regards
Solved! Go to Solution.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО11-26-2001 06:02 AM
тАО11-26-2001 06:02 AM
Re: 12H performance
AutoRAID works in both Raid 0/1 and Raid 5. Raid 0/1 is disk striping with mirroring and Raid 5 is disk striping with distributed parity.
AutoRAID migrates data from Raid 0/1 to Raid 5 and back according to the frequency of use. Raid 0/1 stores the most used data and Raid 5 stores the least used, since Raid 0/1 is much faster but uses more space, whereas Raid 5 is slower but uses space more efficiently.
When it reaches a certain amount, it will continue to migrate from Raid 0/1 to Raid 5 and back all the time, thus reducing efficiency. It is recommended that you leave 50% of the LUN empty so that you have a lot of space allocated for Raid 0/1.
In your case, I think you have to increase an extra hard disk, i.e. 2 harddisks altogether. Also, if you increase one harddisk of higher storage space than the others, it will only allocate the size of the other harddisks in the disk. You have to add two harddisks in order to allocate the full space of higher-capacity disks.
HTH,
Vince
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО11-26-2001 08:59 PM
тАО11-26-2001 08:59 PM
Re: 12H performance
You mean 50% of LUN or 50% of the entire disk array.
Regards
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО11-27-2001 07:01 AM
тАО11-27-2001 07:01 AM
Re: 12H performance
It has to be 50% of the total space of the disk array, to have half the space free for mirroring in Raid 0/1.
HTH,
Vince
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО11-27-2001 09:05 AM
тАО11-27-2001 09:05 AM
Re: 12H performance
Steve
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО11-27-2001 01:11 PM
тАО11-27-2001 01:11 PM
SolutionIf you really want to make your 12H perform, there are two criteria:
1) Each volume group should be comprised of 2 LUN's of equal size. LunA should be primary path Controller X (alternate Y). LunB should be primary path Controller Y (alternate X). You then strip each logical volume in the volume group across both LUNS in 64k stripes.
If all of your Oracle data is going on the 12H, it is perfectly fine to put everything in one Volume Group with multilple LUN's. The important point is that you are now fully utilizing both external scsi paths and thus all four internal paths.
2) Allocate no more than about 60% of the array as LUN's. This keeps the AutoRAID in RAID 10 all the time.
The number of LUN's (other than as outlined above) has no impact on performance and actually 8 LUNS (0-7) are allowed. If you have OnlineJFS, you can use the mounty options convosync=direct,mincache=direct,delaylog,nodatainlog,suid,rw to achieve performance result which are indistinguishable from rawdisk with the convenience of conventional files. Those options bypass the unix file buffers just like raw/io. I would use those options for data and indices and use the options delaylog,suid,rw,nodatainlog for archive and redo logs. If all of this is going to an array, you can actually do this with just two filesystems. DBA's panic but there is no need; the 12H distributes the data across all the disks.
Regards, Clay