section 1 |
|
|
| hp storage white paper |
| |
gives 100X the data protection of traditional | environment changes. HP’s Virtual Array | |||||
RAID 5 and 10X the data protection of RAID 1. | Architecture tunes the array automatically, 24 | |||||
Second: The HP Virtual Array stripes across all | hours per day. It is the | |||||
the disks in the redundancy group. This involves | ||||||
more disks but is not so different from any other | be to compare the multiple manual processes | |||||
RAID 1+0 implementation. Third: array striping | required to set the type, load the paper, and | |||||
is the way of the future. It’s faster and safer. | actually print books on the | |||||
|
|
|
| presses with those of the automated printing | ||
Question: Doesn’t all this “behind the scenes” | systems of our generation. Obviously, automation | |||||
movement of data require a huge performance | in printing presses adds to greater performance. | |||||
hit? Surely, the | It works the same for arrays. | |||||
slow performance. |
|
| Question: Wouldn't striping the data across | |||
Answer: Not true. For years, HP’s original | ||||||
such large RAID groups make the disk rebuild | ||||||
virtual array, | the AutoRAID | Model 12H, | was | |||
times very long and take up a high percentage | ||||||
used for HP’s | benchmark | tests. | ||||
of the array resources? | ||||||
And today, HP Virtual Array products have been | ||||||
Answer: Because of HP’s RAID 5DP this is | ||||||
used for the new rp8400 benchmark testing and | ||||||
are scheduled to be used for the Superdome | practically a | |||||
products. The HP Virtual Array’s RAID 5DP gives | ||||||
mimics the policies of experienced system | each redundancy group 10x the protection of | |||||
administrators. Would an experienced system | RAID 1. RAID 5DP requires that 3 drives would | |||||
administrator do a reconfiguration of the array | have to fail before there would be data loss. | |||||
during a period of high workloads? No! And | Thus, even AFTER a drive fails, the data is STILL | |||||
neither would the HP Virtual Array. The HP Virtual | protected with the same degree of protection as | |||||
Array policy is to NOT perform the background | standard RAID 5. Finally, this means that a single | |||||
tuning operations when the array is under a | drive failure does not put the data at risk and | |||||
greater than 60% workload. |
|
| therefore does not require an emergency | |||
|
|
|
| rebuild. The HP Virtual Array can take the time | ||
Question: Isn’t manual always better than | and do the rebuild in the background without | |||||
automatic? In cars, manual transmissions give | impacting incoming I/Os. Also, the HP Virtual | |||||
better performance than automatic transmissions | Array will rebuild the RAID 1+0 data first since | |||||
because they give the driver more control over | that is the most vulnerable after a failure. In all | |||||
performance; likewise wouldn’t manual RAID | cases, data integrity is ensured and performance | |||||
configurations be better than HP’s Virtual Array | is preserved. No other array can make this claim. | |||||
Technology and its automatic RAID configurations? | Question: If your environment is totally stable, | |||||
Answer: Manual transmissions in cars would | ||||||
wouldn’t a manual configuration by an experienced | ||||||
not give better performance if drivers were only | ||||||
system administrator result in better performance | ||||||
allowed to shift gears once | and could never | |||||
than one derived from the virtual array’s policies? | ||||||
change them after that. In essence, this is what | ||||||
Answer: Let’s first admit it: Totally stable | ||||||
you have with traditional disk arrays. You are stuck | ||||||
with the initial configuration unless you bring the | environments are rare. But the answer is that if | |||||
array down and go through a | the environment were totally stable, and if the | |||||
and complex | reconfiguration every time the | administrator configured the array absolutely |
1.7