Does the number of disks in a RAID-5 array affect the performance ?

Question to the Storage Advisors: What are the practical limitations on the number of disks in a RAID-5 set? I understand that a larger number of disks worsens the probabilities of bad things like a failure or paying the worst possible seek/delay cost in random accesses. Do you have any mathematical models for evaluating these penalties? …

I wanted to let you fine folks know that I’ve recorded a few webinars on many of the topics that I routinely discuss here in this blog. Each one is about 15 minutes long and comes with smorgasbord of tasty slides. The biggest downside is that my delivery is only slightly better than Ben Stein discussing the Hawley-Smoot Tariff Act. (Anyone? Anyone?) …

Question to the Storage Advisors: Which is better: (a) backup battery for cache as found on OEM RAID controllers or (b) writing cache content to one or more disk drives?

Question to the Storage Advisors: I have several large SCSI arrays ranging in size from 600 GB to almost 2 TB. They add up to about 4.5 TB in total. They are currently seen as separate drives in Windows 2003 server R2. I’d like to find an adapter card that will let me span them so they appear as one large array. I’d rather not span them using disk manager in Windows 2003.

Question to the Storage Advisors: The age old question…RAID 5 vs. RAID 0+1 with a twist on spindles. Here’s the deal: Multiple Progress databases. Much more read intensive that write intensive. Which is faster: (2) RAID0 sets of 4 disks that are then mirrored (8 disks total) OR all 8 disks in a RAID5 config? I’m trying to figure out if more disk spindles outperforms less spindles without the RAID5 overhead.

Question to the Storage Advisors: We’re buying 2 cages that use a PCIe interface and 24 146G 15K drives. We’re going to use RAID 10. Each cage only holds 12 drives. We’ll be using 2 spares in each cage. So that’s a stripe of 5 disk groups. At 175 IOPs that’s approx 875 IOPs across the stripe. If we go higher than 875 IOPs we begin seeing higher queue lengths correct? Both cages are hooked to the same server thru the PCIe controller that supports up to 8Gb/s. This is for an Oracle db, and will host 4 dbs. 2 dbs are IO intensive, 2 are not. Right now we plan on separating the 2 IO intensive dbs to 1 cage each. Would it be better to use both cages per db? I don’t think you can stripe disks across cages(?), so would it be better to create two stripes and put datafiles from both dbs on both cages so that we effectively use 2 stripes of 5 disk groups?

Question to the Storage Advisors: How many disk IOs does it take to write one file? I have a RAID10 array writing a 120KB file. Is this one IO for the entire file, or is it one IO per sector on the disk/write cache (which would then be approx 240 IOs)?…

Question to the Storage Advisors: I just wonder, what kind of storage solution that can saturate 10Gb Ethernet backbone. If I assume I have a file server (of course equipped with 10GbE NIC on PCI-E 8x), can an external SAS/SATA-II JBOD box (using 4xwide SAS link) produce enough throughput (taking into account the hard disk speed and the link bandwidth) to saturate the network? If not, where will be the bottleneck?…

Question to the Storage Advisors: I’ve heard great things about SATA based DASD and JBOD devicess. I’ve found tons of information about the data transfer rate, but haven’t been able to find any hard data on how many sustained and bursted IOPS such systems can handle…

Question to the Storage Advisors: For network storage which is better: SATA 150 or SCSI ultra360. Balance size cost and speed. We need a terabyte for live data running and storage…

It’s rare that Digg has a link to a RAID article with enough popularity to hit the front page, but it happened recently with one here from Jon Bach at Puget Systems. I have no idea who Puget Systems is or who Jon Bach is (apparently he’s the president), but the guy made a few interesting points that our readers may find interesting…

Storage benchmarking is what I like to call a “dark art”.

Let’s be honest, the reason File Virtualization is hot is because every compute environment is running into the unavoidable realities of exploding data growth in all directions - spreading like a virus across the network and in the data center.

In the realm of storage appliances (NAS or application specific appliances) the technology evolution usually equates to things like more capacity (bigger drives), faster throughput, more IOPS, better availability, and easier management/provisioning, etc… but is that it?

We’ve recently had some questions coming through the Ask the Storage Advisors link about networked storage in the home.

can a RAID 1 be setup between the two servers creating a real-time duplicate of the active server on the backup server?

Can you use software RAID5 across multiple hardware RAID5 or RAID0 storage systems using iscsi?

Question to the Storage Advisors: I now read that there are specialized RAID hard disks that have “Time Limited Error Recovery”, which my disks do not have. Will my RAID controller compensate for hard disk based error recovery that takes many minutes?

Question to the Storage Advisors: Why is such a large drive buffer/cache needed, considering controllers and the host have much larger caches anyway? And/or, why is read-ahead data not forwarded to the host immediately?

Question to the Storage Advisors: I’m torn between RAID 5 and 10. I know that with RAID-5, the performance of a third-party card is far superior than using an on-motherboard controller, but what about RAID-10? …

Question to the Storage Advisors: I am looking for advice on how to calculate throughput (MB/sec) from an IOPS value…

Question to the Storage Advisors: Tom’s Hardware recently did a report on the newer P965 motherboards with Intel’s new ICH8R Southbridge which supports their Intel Matrix RAID technology where you can have more two RAIDs on the same 4 drive set…

Let’s face it… finding the “right” data is hard, but there is an answer.

If some of the new indexing applications live up to the hype …. ILM may actually have a chance…

Question to the Storage Advisors: I have a question regarding RAID. With dual processor CPUs now common (e.g., Pentium D) wouldn’t a 4-drive RAID-5 array be faster than a 4-drive RAID-10 array? I would think the second core would (could?) be dedicated to parity calculations…

Question to the Storage Advisors: We are trying to build a data streaming system for data acquisition purposes. … What would be the best Adaptec controller to use for sustained writing at 100 MB/s for at least 500 GB of data?

A few days ago I sat through an SNW session presented by the venerable Scottsman Willis Whittington of Seagate…

Steve, Joe and I just finished filming a short segment with the effervescent Jon Toigo for his StorageTV website.

This new standard is “FC Base-T”. The proposal came out last August 2005, but I’d never heard of it - perhaps I live under a rock.

I really try to avoid blogs-about-blogs (BAB - I just made that up), but there was an interesting post over on DrunkenData. There is some odd disagreement on that site about whether NetApp is cheating their customers by wasting a bunch of physical storage…

What is your definition of CDP! Pick one that works for your business!

What data storage solutions are available for agriculture?

Yesterday I wrote briefly about the use of RAID in the home. I think we can all agree that the home contains data worthy of being protected against drive failure. Now it’s just a matter of figuring out how to get ‘er done.

Is there a need for RAID in the home, or am I just a guy with a hammer looking for a nail?

Question to the Storage Advisors: I am trying to help an IT manager friend of mine to understand which solution would be the best one to store a LARGE amount of very small files.

Question to the Storage Advisors: While most of us are more or less clear on calculating disk IOPS, is there a similar way of calculating cache IOPS?

Network World posted an article last week called The High Price of RAID-5, written by Mike Karp….

Question: Regarding SATA vs SATA II for an older PC Workstation, will using an add-on SATA II controller really help the data read/writes to a RAID 1 pair of disks if my system bus is a PCI 33MHZ bus? Would a SATA 150 controller do just as good? In other words, where is the bottleneck?…

Question: With SAS is spindle speed as important as was with older technologies? For example, six 15K RPM drives vs eight 10K RPM — which would perform faster?…

It appears that the Great Storage Blog Experiment has been a success. One way I know is by looking at the number of product-specific inquires we’re getting through the “Ask the Storage Advisors” link…

From my admittedly jaded point of view it feels like the storage industry has done plenty to promote SAS as the replacement for SCSI. I’ve certainly posted on the topic enough, and I’ve given several presentations and webinars touting the virtues of SAS. My guess is that all the enterprise folks that use SCSI or FC now certainly know about SAS.

But it feels like the channel doesn’t get it yet….

Question: Is there a rule of thumb concerning RAID-5 Block Stripe Size to file size? Is the any direct performance correlation between Block Strip Size and NTFS cluster size?…

Why should you separate RAID for the OS from RAID for your data?

I want to give a quick shout-out to my home-boy Jon Toigo for an excellent event over in Tampa this week. It was his first, and hopefully not his last, Disaster Recovery and Data Protection Summit…

Adaptec wins Oscar…

We often have very good comments on posts go unnoticed. So I’m making this post to point out a good comment from Allen regarding dual-ported SAS.

Question: Is a StorageTek Iceberg 9200 worth anything or is it outdated?

The folks over at Tom’s Hardware (no relationship) have recently posted an article on SAS…

Question: I just “suffered” through another webinar touting the virtues of SATA without really going into the MTBF or MTBDL (data loss)…

A few weeks ago I spoke at a TechData Government Conference here in Orlando. The title of my talk was “Serial Attached SCSI: Get Started Today”. I had a pretty good crowd of ~25 people…

MAID stands for Massive Arrays of Idle Disks. It was first mentioned in the January 2002 whitepaper of the same name, created by Colarelli, Grunwald and Neufeld from the University of Colorado in Boulder…

Question: In the late 90’s there was a firestorm of controversy surrounding the lifespan of CDs (to which Tom recently alluded). Is there a definitive answer? Is it safe to use CD-Rs as a permanent archive?

The problem is all in the math. Most of us that have been in the storage industry know that the way the hard drive vendors “market” the drive capacities is fundamentally inaccurate.

I recently was asked about iFPC and FCIP as competing
> technologies

…Unlike pressed original CDs, burned CDs have a relatively short life span of between two to five years, depending on the quality of the CD…

Question: Do you think the stability and reliability of SATA drives will improve? We have not had a lot of luck with them in production and the overall opinion is that ATA is a desktop technology vs SCSI which is designed for the server end.

Question: Where do FATA drives fit in?

Question: Is there a physical difference in edge and fanout expanders, or is it just a matter of where they exist in the SAS domain? In other words, if you surround an expander with edge expanders is that central expander considered a fanout?

Question: Are there any constraints on the numbers of SAS initiators and targets you can mix & match?

Question: Is InfiniBand pronounced with a long “e” as in “IN fin e BAND”?

Question: I haven’t been able to figure out who developed RAID 6 and how long it’s been around. Any ideas?

Question: I heard that the external SAS cables will have direction orientation. Can you explain why and will there be markings to let you know which way to go?

A quick word about a webinar that Steve Rogers and I gave last week…

All RAID-6 is not alike. There are pros and cons to each, but they all support two simultaneous drive failures. …

These backup services are quite convenient. It does require an Internet connection, which just about everyone has now.

…in my opinion this is a prime example of people being lazy or stupid!

The toughest question for any storage administrator to answer is, “What kind of performance do you NEED?”

There’s been lots of talk about RAID 6 as it compares to RAID 5 and the reliability benefits that are gained by SATA drives when using RAID 6 – so I’ve decided NOT to talk about that anymore!
Instead I thought I’d discuss two other RAID levels out there and available to folks. RAID 5EE […]

…I profess that, unequivocally, iSCSI has indeed changed the face of storage networking - forever!

Finally, a solution for low-volatility storage enclosures suffering from structural framework issues. …

For a long time now I’ve been a proponent of replacing tape drives with hard drives. But there have always been technical hurdles to overcome…

… how bad would SATA have been without the benefit of RAID-6? …

SAS ready for the Enterprise? Just how Big? and what about HA?

I’m not going to go into the values of having applications and services available over the web – I think that’s a personal preference for each individual and company to have to decide.

Storage Management in Rough Waters

Now that we know how to calculate the Mean Time To Data Loss (MTTDL) for RAID-5 and RAID-6 arrays, let’s take a look at some real life cases using actual SATA and SAS/SCSI arrays. …

It’s true I tell you – you can absolutely deploy SAS and SATA drives within the same physical enclosure… but be careful!

There are a number of ways that IT Pros deal with remote site data protection – some are good and some are not so good…

In an earlier post Joe mentioned that the reliability of SATA could be improved by using RAID-6. That’s very true, and there’s a good story to be told about the reliability of SATA with RAID-6 being much higher than the reliability of SCSI/SAS with RAID-5. So I thought I would start by reviewing how reliability is determined for both types of arrays. …

Finally these “third class citizens” (SATA drives) can get a little respect with a little help from RAID 6.

Byte and Switch poll: Is SAS a viable alternative to Fibre Channel as an enterprise-class drive?

SAS, RAID, Virtualization & Storage Management

Storage is expensive because of all the crazy custom hardware. But you don’t need that. The value is in the software. Spending money on fancy sheet metal isn’t going to make your data any safer.

Who has your data?

Storage Managment at the Show

SAS will take major market share from FC, some from ATA/SATA, and even share from iSCSI.

I’ve heard many questions lately about SAS (Serial Attached SCSI). Simple questions like, “What’s SAS?” or, “Why should I consider SAS?” This information can be found with simple web searches. Adaptec has a good section on the website dedicated to the education of SAS and will answer these questions easily and the […]

After all of the efforts over the years to agree on a standard with really no end result that the end users can see, hear, feel and touch – how could we really think that a standard could ever possibly prevail?

…an IT Manager asked me what I thought of 10 Gigabit Ethernet and it’s effects on Fibre Channel.

… Microsoft has been putting out feelers to customers and vendors over the last several months regarding dropping support in the Vista timeframe for PCI RAID clustering. …

It appears that Wilma has claimed its first SNW victim - Intel. I just received a message this morning saying that “due to Hurricane Wilma, Intel has decided to cancel our participation in next week’s Storage Networking World in Orlando, Florida”. …

In an effort to insight a riot with one of my fellow Storage Advisors I want to talk about “Virtual” Tape Library (VTL) options.