10-17-2012
Well, I am sure most SAN can be configured for mirroring, which is raid-2 or something like that. To achieve the most bang for your buck, you want the mirrors to be as far apart as possible, so something between the application host and short of the storage box needs to know where there are 2 so it can sidestep the dead side as well as splitting the query load and duplicating the churn. The farther upstream it is done, the greater the reliability, but too close can load the app server and communications, unless fiber has multicast write and anycast read. There may be several layers vying to mirror your storage, so pick wisely.
Mirroring got overshadowed a bit by raid, but has always had a query bandwidth advantage, with two devices handling query. Within the devices, there can be as much striping as in raid, so that is no different. When writing, there is no parity calculation and additional write time, just two immediate simultaneous writes. With raid in sequential striping mode, you write data to 1,2,3,4 and parity to 5, then data to 5,1,2,3 and parity to 4, and so on, so while reading is 5x spindle speed, writing is 4x. A mirrored pair trades space for bandwidth. Disk is cheap, and bandwidth is golden. Finally, it seems some raid systems seem to only get defects detected by staff when 2 adjacent devices fail, so often raid5 is either also mirrored or great downtime, data loss and partial restore pain is experienced.
10 More Discussions You Might Find Interesting
1. Solaris
Hi everyone,
I wonder if I can canvas any opinions or thoughts (good or bad) on SAN attaching a SUN V880/490 to an EMC Clarion SAN?
At the moment the 880 is using 12 internal FC-AL disks as a db server and seems to be doing a pretty good job. It is not I/O, CPU or Memory constrained and the... (2 Replies)
Discussion started by: si_linux
2 Replies
2. AIX
Hi guys,
I'd like to share my migration/mirroring of ssa to san. No downtime for users, probably I/O performance.
here's the step:
1 After the lun had been carved on the SAN and the connections had been done on AIX fiber card
2 “lspv” and look for the new SAN hdisk? on the bottom, say... (1 Reply)
Discussion started by: itik
1 Replies
3. AIX
Hi All,
I would like to share this incident that happened the other day.
I have a question with this, https://www.unix.com/aix/64921-create-new-vg-san-rename-fs.html
And I thought it's related to the above link but the problem was the ibm san 4300 cache battery was dead and I need to click... (2 Replies)
Discussion started by: itik
2 Replies
4. Filesystems, Disks and Memory
Hello everyone !
Im new on Ibm San DS4500.
Can you give me some tips to this, because I dont want to make a mistake.
I have some questions.
How can I know how much space get on the san, I cant find it.
How can add more space to a partition.
Do you have some tutorial about this. I... (0 Replies)
Discussion started by: lo-lp-kl
0 Replies
5. AIX
I have a question about SAN commands
I have almost 15Tb of disk on my san
but assigned and something else I have almost 11Tb
There is a command to know, what its my real total storage capacity and
another command to know how much I used .?
Thanks again in advance (0 Replies)
Discussion started by: lo-lp-kl
0 Replies
6. AIX
Hello,
I have AIX 6.1 with TL 4 and it is connected to IBM SAN STORAGE DS4700
After assigning some disks from SAN to AIX, I can see the disks in my AIX as
hdisk2 Available 05-00-02 MPIO Other DS4K Array Disk
hdisk3 Available 05-00-02 MPIO Other DS4K Array Disk
But it should... (0 Replies)
Discussion started by: filosophizer
0 Replies
7. Solaris
HI all,
I had recently change the Server storage from EMC to the IBM SAN.
but after the configuration, the IBM success to see the server HBA port and successfully assign a LUN for the server.
When i go to the server, and restarted it. i use the "format" command to check, but din see any... (1 Reply)
Discussion started by: SmartAntz
1 Replies
8. AIX
Hello,
I have IBM SAN STORAGE DS4100 and one of the cache battery for the controller is dead. Suddenly the performance has been degraded and access to SAN disks ( reading and writing ) became very slow ?
My query: Replacing the battery will take 6 days, so in the mean time what are the ways... (1 Reply)
Discussion started by: filosophizer
1 Replies
9. AIX
Hello,
I have DS4000 IBM SAN Storage ( aka FastT Storage )
One of my disks has failed and I had a hot spare disk covering all the arrays. As the disk failed, immediately the hotspare disk took over the failed disk ( see the JPEG in the attachment )
My Question: How can I make the hotspare... (1 Reply)
Discussion started by: filosophizer
1 Replies
10. AIX
Hi,
This is follow up to the post https://www.unix.com/aix/233361-san-disk-appearing-double-aix.html
When I connected Pseries Machine HBA Card ( Dual Port ) directly to the SAN Storage DS4300 , I was able to see Host Port Adapter WWN numbers , although I was getting this message... (2 Replies)
Discussion started by: filosophizer
2 Replies
LEARN ABOUT REDHAT
raidtab
raidtab(5) File Formats Manual raidtab(5)
NAME
raidtab - configuration file for md (RAID) devices
DESCRIPTION
/etc/raidtab is the default configuration file for the raid tools (raidstart and company). It defines how RAID devices are configured on a
system.
FORMAT
/etc/raidtab has multiple sections, one for each md device which is being configured. Each section begins with the raiddev keyword. The
order of items in the file is important. Later raiddev entries can use earlier ones (which allows RAID-10, for example), and the parsing
code isn't overly bright, so be sure to follow the ordering in this man page for best results.
Here's a sample md configuration file:
#
# sample raiddev configuration file
# 'old' RAID0 array created with mdtools.
#
raiddev /dev/md0
raid-level 0
nr-raid-disks 2
persistent-superblock 0
chunk-size 8
device /dev/hda1
raid-disk 0
device /dev/hdb1
raid-disk 1
raiddev /dev/md1
raid-level 5
nr-raid-disks 3
nr-spare-disks 1
persistent-superblock 1
parity-algorithm left-symmetric
device /dev/sda1
raid-disk 0
device /dev/sdb1
raid-disk 1
device /dev/sdc1
raid-disk 2
device /dev/sdd1
spare-disk 0
Here is more information on the directives which are in raid configuration files; the options are listen in this file in the same order
they should appear in the actual configuration file.
raiddev device
This introduces the configuration section for the stated device.
nr-raid-disks count
Number of raid devices in the array; there should be count raid-disk entries later in the file. (current maximum limit for RAID
devices -including spares- is 12 disks. This limit is already extended to 256 disks in experimental patches.)
nr-spare-disks count
Number of spare devices in the array; there should be count spare-disk entries later in the file. Spare disks may only be used with
RAID4 and RAID5, and allow the kernel to automatically build new RAID disks as needed. It is also possible to add/remove spares run-
time via raidhotadd/raidhotremove, care has to be taken that the /etc/raidtab configuration exactly follows the actual configuration
of the array. (raidhotadd/raidhotremove does not change the configuration file)
persistent-superblock 0/1
newly created RAID arrays should use a persistent superblock. A persistent superblock is a small disk area allocated at the end of
each RAID device, this helps the kernel to safely detect RAID devices even if disks have been moved between SCSI controllers. It
can be used for RAID0/LINEAR arrays too, to protect against accidental disk mixups. (the kernel will either correctly reorder disks,
or will refuse to start up an array if something has happened to any member disk. Of course for the 'fail-safe' RAID variants
(RAID1/RAID5) spares are activated if any disk fails.)
Every member disk/partition/device has a superblock, which carries all information necessary to start up the whole array. (for
autodetection to work all the 'member' RAID partitions should be marked type 0xfd via fdisk) The superblock is not visible in the
final RAID array and cannot be destroyed accidentally through usage of the md device files, all RAID data content is available for
filesystem use.
parity-algorithm which
The parity-algorithm to use with RAID5. It must be one of left-asymmetric, right-asymmetric, left-symmetric, or right-symmetric.
left-symmetric is the one that offers maximum performance on typical disks with rotating platters.
chunk-size size
Sets the stripe size to size kilobytes. Has to be a power of 2 and has a compilation-time maximum of 4M. (MAX_CHUNK_SIZE in the ker-
nel driver) typical values are anything from 4k to 128k, the best value should be determined by experimenting on a given array, alot
depends on the SCSI and disk configuration.
device devpath
Adds the device devpath to the list of devices which comprise the raid system. Note that this command must be followed by one of
raid-disk, spare-disk, or parity-disk. Also note that it's possible to recursively define RAID arrays, ie. to set up a RAID5 array
of RAID5 arrays. (thus achieving two-disk failure protection, at the price of more disk space spent on RAID5 checksum blocks)
raid-disk index
The most recently defined device is inserted at position index in the raid array.
spare-disk index
The most recently defined device is inserted at position index in the spare disk array.
parity-disk index
The most recently defined device is moved to the end of the raid array, which forces it to be used for parity.
failed-disk index
The most recently defined device is inserted at position index in the raid array as a failed device. This allows you to create raid
1/4/5 devices in degraded mode - useful for installation. Don't use the smallest device in an array for this, put this after the
raid-disk definitions!
NOTES
The raidtools are derived from the md-tools and raidtools packages, which were originally written by Marc Zyngier, Miguel de Icaza, Gadi
Oxman, Bradley Ward Allen, and Ingo Molnar.
SEE ALSO
raidstart(8), raid0run(8), mkraid(8), raidstop(8)
raidtab(5)