Reliability Ranks #1 As Most Important Factor In Enterprise Storage Purchasing

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
According to Western Digital Corporation’s 2nd annual global survey of CIOs and IT decision makers, reliability trumps cost when it comes to making data-storage purchasing decisions as organizations leverage data to drive growth, profitability and shareholder value. According to the survey, reliability of data storage systems outranks cost as the most important strategic consideration for 2016, indicating how price sensitivity has given way to the importance of protecting and gleaning insight from data.

A full 85% of the respondents recognize the positive impact data can have on their bottom line, but are playing catch-up as 55% admit they are not yet storing all the data necessary for long-term business success. To address this, 74% of the IT decision makers are growing cold storage and archiving capacity, while 81% are investing in data analytics to prioritize infrastructure investments and mine internal data assets. The survey also shows that cloud infrastructure budget and build-out initiatives have increased in order to better store and access the growing trove of data.
 
Steve said:
According to Western Digital Corporation’s 2nd annual global survey of CIOs and IT decision makers, reliability trumps cost when it comes to making data-storage purchasing decisions
Click to expand...

Only to a limited extent. I'm not going to pay 3 times the price for 10% better reliability.

Higher level raids can also make up for slightly lower reliability.
Using Raid 6, 10, 50 or 60, depending on the size, and using a weekly data scrubbing on the raid can make up for slightly less reliability.
 
Last I checked reliability took a backseat to price. But that's just every company I've ever worked for so what do I know.
 
Last edited:
nutzo said:
Only to a limited extent. I'm not going to pay 3 times the price for 10% better reliability.

Higher level raids can also make up for slightly lower reliability.
Using Raid 6, 10, 50 or 60, depending on the size, and using a weekly data scrubbing on the raid can make up for slightly less reliability.
Click to expand...

May you would. 10% better reliability @ 3x the cost MIGHT pay for itself compared to associated downtime, list revenue, etc.

I know my exact burden rate at work. For example, at my current rate, if I ever made a patch cable instead ordering one and have it shipped first AM delivery..I would be an idiot. Yes I just paid $60 dollars for delivery on a $2 cable and I'm ahead of the game!
 
I agree with Western Digital.
.
.
.
.
.
.
.
So buy HGST drives. :D :D
 
bomni said:
Last I checked reliability took a backseat to price. But that's just every company I've ever worked for so what do I know.
Click to expand...

My company went through three different SAN solutions before the downtime taught them to pony up a few bucks and get something that will work.
 
We support an electronic medical record used by nursing homes and assisted living - something in the range of 1200 residents in 10 locations. Downtime is BAD. Downtime means your grandmother has to wait longer than she might otherwise have to because you can't pass pain meds without checking the EMR.

So, we pay through the nose for EMC.

I whitebox our cold storage and the cold storage backup. Those servers aren't even in the DC.
 
The fact that Seagate is even still a company blows the OP out of the water.
 
bomni said:
Last I checked reliability took a backseat to price. But that's just every company I've ever worked for so what do I know.
Click to expand...

Yeah the problem is this is just a survey of IT people, who of course are going to say reliability is important because they know what they are talking about. When you start factoring in cost and trying to justifying those costs to management things change in a hurry. Cutting corners for short term gains is standard practice in most modern companies.
 
Last edited:
As a former PC Tech and now sysadmin (25 years collectively), I'd pay up to 50% more for something that was 10% more reliable. Mission critical uptime is far more important than performance.

fyi I've seen every brand of HD experience widespread reliability problems.
 
Makes sense, time is money so downtime means down money including the labor to fix the issue.
 
Not just more reliability with less downtime, but the level of support to get you back up VERY quickly in the event of a catastrophic failure. 4 hours vs NBD.
 
ecktt said:
As a former PC Tech and now sysadmin (25 years collectively), I'd pay up to 50% more for something that was 10% more reliable. Mission critical uptime is far more important than performance.

fyi I've seen every brand of HD experience widespread reliability problems.
Click to expand...

For shops running their own a small amount of infrastructure (<100 nodes), definitely helps to have more reliable drives. For the enterprises running larger infrastructures, they will usually mature their applications to live on ephemeral hardware so that drive failures are non-events that get taken care of during maintenance windows.

It's even easier now for smaller shops utilizing services like Amazon's S3. Configurations get shipped into S3, and the local stuff is made ephemeral and can be automated to be built off of the "cloud".

For my employer we have several thousand machines and they go down here and there. They are failed over to a backup instance during the day and the datacenter guys swap in a new box or replace the disk and we point back to the primary instance.
 
CombatChrisNC said:
Not just more reliability with less downtime, but the level of support to get you back up VERY quickly in the event of a catastrophic failure. 4 hours vs NBD.
Click to expand...

I've had the 4 hour replacement with Dell and IMO, it's a PITA. Not sure about others, but by the time one has troubleshooted the issue, then determined that replacement is necessary, getting through calling and opening a ticket then troubleshooting again, then escalating enough to get the 4 hour work order placed, it's been 6-10 hours.

It doesn't make sense to me that an app can go down for just ~4 hours and be okay. It just makes sense to have standby/backup hardware on hand for critical components of a system and replace bad hardware at ones leisure.
 
You must log in or register to reply here.
Back
Top