Hard drive replacement programs

sec4884 · Oct 3, 2016

Hi All,

What does this group think of hard drive replacement programs? We have arguments for replacing only after failure vs. attempting to put in a preventative maintenance plan. This would be for panel mounted PCs running iFix / Wonderware / vendor specific HMI software.

Also what does this group think about hard drive imaging options for disaster recovery?

Thanks in advance for fueling our argumentative fire over here!
Stacey

fredz0003 · Oct 3, 2016

I would say is a waste of money to replace a good drive just for the sake of preventative maintenance. Take full images of the hdd regularly of any time there's a major update. Then if hdd fails just restore the image. I do this a lot at work we use imagex, but you could use acronis too. With imagex process takes less than 30 min depending on size.

Edit: if you have super crucial systems / servers use raid for sure.

Crammed77 · Oct 3, 2016

Hey Stacey,

At my plant, we replace hard drives after failure. While we have PM's in place for just about everything else, I do not get overly excited about the thought of replacing hard drives just because a time is up.

We protect ourselves other ways.. We use Symantec System Recovery on our critical PCs, and this has been enough to both acknowledge potential failures and get us up and running quickly if a hard drive or PC does fall over.

We also have some of our process PCs set up in a RAID configuration, which to be honest has been both good and bad really. The other thing that we have out here are PCs that can be swapped over in case of emergency.

sec4884 · Oct 3, 2016

We are likely going to Commvault but currently have a bunch of acronis images just sitting on an external hard drive. And people thinking we should start replacing hard drives based off of mean time to failure (agree this is silly).

Out of curiosity what have been the bad things coming out of RAID configuration? Seems like a good option if our AB computers actually allowed for such a thing.

fredz0003 · Oct 3, 2016

I can't remember our raid setup but in one specific occasion the raid kept degrading because one drive wasn't like the other ones and the raid manager didn't like that. They do tend to be painful when issues arise.

In theory it should as easy as swapping the bad drive and letting the manager rebuild the raid on the background.

Crammed77 · Oct 3, 2016

fredz0003 said:
I can't remember our raid setup but in one specific occasion the raid kept degrading because one drive wasn't like the other ones and the raid manager didn't like that. They do tend to be painful when issues arise.

In theory it should as easy as swapping the bad drive and letting the manager rebuild the raid on the background.

This was a definite problem for us, the other thing that I seem to remember was that some were set up from BIOS, others were set up through Windows, the time it takes to rebuild the RAID array..

I am sure that there are better IT ways of doing it than I was doing, but when the plant is down, the past thing that I want to do is try to finesse a RAID controller back to life when it is much easier to just throw a new hard drive in.

Epy · Oct 3, 2016

Basically comes down to this:

Q: Do you back up everything important and have OS images you can plop onto a new hard drive with ease, minimizing downtime?

If yes, then replace HDDs when they fail. If not, replace early.

Modern HDDs have S.M.A.R.T. that aid in failure prediction. Get distributed software that monitors the SMART statuses, replace on a decided threshold.

Jeebs · Oct 4, 2016

We have had so many RAID related issues, we ditched the concept entirely.
We now have 2 HDD's, one plugged in and in use, the other not plugged in. These are drives in hotswap bays so you just turn the power off, pull the defective drive, push the back-up in and reboot. Up and running in minutes. Recreate back-up with offline image and place drive in bay, ready to be used on a moments notice.
The best part? The swap can be done by even your least accomplished sparky. So no need to come in on a Sunday morning at 3am just to replace a defective HDD.

rdrast · Oct 4, 2016

Replace on failure.
We keep 3 images per week of each HMI machine in the plant on our server, so redeploying isn't a big deal if necessary.

An even better approach would be to use a VM on every floor PC, and just keep the snapshots updated, but I haven't gone that far yet.

splicer480 · Oct 4, 2016

I backup my drives using Acronis.
I've recently swapped out most of our 2.5" standard drives with an adapter board bought through Newegg. It allows me to use a 266x Compact Flash card as a hard drive. Just make sure it is a Kingston brand CF card because they are bootable.
The main problem we were having was operators shoving the arm mounted HMI/Computer combo too hard and when it hit it's stop, sometimes it would scratch the (spinning) hard drive destroying it.
I don't have to worry about that now going solid state.

isiah_s · Oct 4, 2016

In data centers the policy is to replace the entire raid array as a unit. To avoid the situation where drivers of different types get mixed. So, if you had a set of drives raided with a 5 year warranty. In year 6 you could look at replacing the whole set at once. Drives like all physical things show a bathtub curve of failures. Meaning your failures are most likely to show up in the beginning and end.

Hard drive replacement programs

sec4884

Member

fredz0003

Member

Crammed77

Member

sec4884

Member

fredz0003

Member

Crammed77

Member

Epy

Lifetime Supporting Member

Jeebs

Member

rdrast

Lifetime Supporting Member

splicer480

Member

isiah_s

Member

Similar Topics