Comments on system build?

[kony]

That's a term from danabase management. In this context it means
restoring the last backup.

Yes I know, I was just questioning it as a real strategy to
do anything more than uninstall a bad driver- something that
doesn't necesssarily require a rollback at all.

I do not understand you.

There isn't anything all that positive about a rollback, why
do it? Just turn off system restore and make the regular
backups, since an entire backup certainly covers and gain
you might see from the rollback, plus a lot more coverage.

Please explain the use of the term comprehensive in this context.

Covers more scenarios. Could be a failing drive, a virus, a
bad driver, system instability, etc, etc. The off-drive
backup covers all these instead of several different
strategies that only target a few things (potential
problems) each.

Displace?

Yes, there is no need for a rollback if one has a good
backup strategy. Particularly important might be to keep
backups of OS partition and documents separate, so that your
old docs aren't overwritten. This works nicely into make
backup strategies since the documents are going to change
more often anyway, at least once a system is finalized and
fully configured, in it's everyday-running state.

I am talking about keeping a complete hardware backup of the HD each
morning at 4:00 am. That way if something goes wrong during the day,
depending on the extent of the problem, you can recover the new data
onto a 3rd disk (eg. use xcopy /D and then replace the corrupted HD
with the backup version you made earlier that morning.

That's plenty comprehensive and displaces nothing I am aware of.

Oh, then we have a different definition of rollback. More
often I hear the term used for a driver rollback or misused
to mean a system restore point.

 

[kony]

I am talking about reliable drives like WD. I cannot vouch for crap
drives. I am sure you have seen just about everything. But since I use
quality drives, I have never had a problem.

I am too. WD is certainly included in those that'll die
with no prior warning. FWIW, WD aren't necessarily any
better (on average) than the other big 3 or 4. Just ask
anyone who's had a WD fail, they'll tell you what crap they
are, or someone who had Maxtor fail, then Maxtors are crap.
Bottom line- OEMs have a good grasp of failures and use all
the major brands.

I used to have problems with Seagate crap so I quit buying it and
switched to WD. In approx. 15 years I have never had a WD drive cause
me any trouble.

TYpically the failure rates run around 5% (IIRC), so of
course you'd have to go through 20 of 'em to expect a
failure within their expected lifespans.

Or do you propose the OP buy crap hardware and then shore it up with
RAID 1.

?
Who said WD is expensive? Same low-cost drives on sale and
rebates every other week.

I did not use the term ECC. I use the term error correction, as in
RAID with error correction. I do not recall which RAID it is, but it's
the one where you have enough discs to keep extra information to
correct errors on the fly. Maybe it's RAID 5.

Yes I see what you mean, yes but that's even more drives.
While it's a good option for some people, for others their
system case simply won't keep several drives cooled very
well.

What is not that expensice or difficult to implement? RAID 1?

Many boards support it now, or a controller for under $20.
A 2nd drive, $40, or more if you want larger. Having ~100GB
of RAID1 for under $100 is quite a bargain.

That is
not the issue. The question I am raising is whether it is even needed
if you buy quality HDs to begin with and keep a daily/weekly set of
complete backup disks.

You're not buying insurance when you buy a WD. They're fine
drives, but not bulletproof. Whether a daily/weekly backup
is sufficient depends on actually doing that backup and the
value of accumulated data (and/or time) inbetween each
backup. For some people, that's worth $40-100.

 

[Bob]

Yes I know, I was just questioning it as a real strategy to
do anything more than uninstall a bad driver- something that
doesn't necesssarily require a rollback at all.

Of course it all depends on the extent of the damage. But then there
is the consideration of what your time is worth. It is so easy to
capture the latest data using any number of tooks - xcopy works just
fine for most situations - and then swap the drives. At most that's a
10 min procedure and as long as you did not lose anything of
significance that xcopy could not capture then you are guaranteed that
the system has been fully restored.

How many times have you installed a bad app, spent a lot of time
trying to get rid of it only to find later that remnants are still
present? By implementing a h/w backup, you know the contamination has
not propagated.

There isn't anything all that positive about a rollback, why
do it?

See above.

Just turn off system restore and make the regular
backups, since an entire backup certainly covers and gain
you might see from the rollback, plus a lot more coverage.

Once again you have completely lost me in your terse jargon. I do not
know what "turn off system restore" means. And of course I make
regular backups onto another HD. Your composition appears incoherent -
it must be late.

There is more than just a rollback involved. If important data has
arrived on your HD since the last h/w backup, then you need to capture
it with something like xcopy /d: onto a 3rd disk. Then after rolling
back to the backup disk, you can bring it up to date with the data on
the 3rd disk that you captured from the corrupted disk.

Covers more scenarios. Could be a failing drive, a virus, a
bad driver, system instability, etc, etc. The off-drive
backup covers all these instead of several different
strategies that only target a few things (potential
problems) each.

You have lost me. I am talking about an off-drive backup. The backup
scheme I am proposing requires a 2nd HD to make the backup to each
day.

Yes, there is no need for a rollback if one has a good
backup strategy.

I do have a good backup strategy.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[Bob]

I am too. WD is certainly included in those that'll die
with no prior warning. FWIW, WD aren't necessarily any
better (on average) than the other big 3 or 4. Just ask
anyone who's had a WD fail, they'll tell you what crap they
are, or someone who had Maxtor fail, then Maxtors are crap.
Bottom line- OEMs have a good grasp of failures and use all
the major brands.

I do not count those failures where the user did not treat the drive
with respect. You cannot fault WD just because some idiot fries one of
their drives.

How many of the people who claim WD is junk can tell you what the
operating temperature was? I bet zero.

TYpically the failure rates run around 5% (IIRC), so of
course you'd have to go through 20 of 'em to expect a
failure within their expected lifespans.

Bad thinking. Those 5% are most likely due to poor treatment.

I could go thru 100 ED drives and very likely not find one that is
bad. Of course, I get to specify the vendor, so I know some gorillas
working for a poor vendor didn't play basketball with the drive.

I think if you took the time to chase down the WD drive failures you
have seen, that in essentially every instance it will be the result of
improper treatment.

Who said WD is expensive? Same low-cost drives on sale and
rebates every other week.

I have seen some real junk for half price. Just look at Worst Buy's
weekly fliers.

Yes I see what you mean, yes but that's even more drives.
While it's a good option for some people, for others their
system case simply won't keep several drives cooled very
well.

Then they need to get a server tower.

Many boards support it now, or a controller for under $20.
A 2nd drive, $40, or more if you want larger. Having ~100GB
of RAID1 for under $100 is quite a bargain.

If it is useful. Otherwise it is just a waste of money.

You're not buying insurance when you buy a WD. They're fine
drives, but not bulletproof.

That is correct. You cannot put bullets into them and expect them to
work. You have to keep them cool, you have to make sure they do not
get dropped. You can't move them around while spinning (or you
shouldn't), etc.

But if you treat a WD drive with proper care, you can expect a good
drive. The few that die will do so in the infant mortality stage,
which is something any h/w item can suffer from. So, you put one in on
a spare channel and burn it in for a few days. SOP.


Whether a daily/weekly backup

is sufficient depends on actually doing that backup and the
value of accumulated data (and/or time) inbetween each
backup. For some people, that's worth $40-100.

If the data must be current, then you have no other choice but to
implement RAID. But most people do not have that problem so RAID is a
waste of money. They should spend the money they save on a good case
and several good fans.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[kony]

Of course it all depends on the extent of the damage. But then there
is the consideration of what your time is worth. It is so easy to
capture the latest data using any number of tooks - xcopy works just
fine for most situations - and then swap the drives. At most that's a
10 min procedure and as long as you did not lose anything of
significance that xcopy could not capture then you are guaranteed that
the system has been fully restored.

XCOPY meaning the DOS command, right? That's fine for some
but remember that WinXP defaults (and is, on OEM boxes)
running in NTFS, not FAT32.

How many times have you installed a bad app, spent a lot of time
trying to get rid of it only to find later that remnants are still
present? By implementing a h/w backup, you know the contamination has
not propagated.
Exactly.



See above.

We were using a different definition of rollback.

Once again you have completely lost me in your terse jargon. I do not
know what "turn off system restore" means. And of course I make
regular backups onto another HD. Your composition appears incoherent -
it must be late.

What OS are you talking about if you're not familiar with
"system restore"?

There is more than just a rollback involved. If important data has
arrived on your HD since the last h/w backup, then you need to capture
it with something like xcopy /d: onto a 3rd disk. Then after rolling
back to the backup disk, you can bring it up to date with the data on
the 3rd disk that you captured from the corrupted disk.


You have lost me. I am talking about an off-drive backup. The backup
scheme I am proposing requires a 2nd HD to make the backup to each
day.


I do have a good backup strategy.

Again, you are using a common and "official" Windows term,
"rollback" to describe something else. I'm not suggesting
it's a bad term to use based on the context, BUT,
unfortuately when MS usurps some phrase for their own
purposes, it then becomes the common definition since the
majority of PCs run windows.

 

[kony]

I do not count those failures where the user did not treat the drive
with respect. You cannot fault WD just because some idiot fries one of
their drives.

Neither do I. Put a handfull of drives in a rack, all with
good power and cooling, and sometimes it's the WDs that
fail, not the other brands. If WD were the only brand that
didn't fail, the others would be out of business already.

How many of the people who claim WD is junk can tell you what the
operating temperature was? I bet zero.

We don't have to care, because most people run a
pre-configured OEM box that either had a WD in it or some
other brand. They don't change their system and whichever
drive was installed, could fail. In that most common
scenario, WD drives fail too, at similar rates.

Bad thinking. Those 5% are most likely due to poor treatment.

It would be ridiculous to claim that the "poor treatment"
doesn't extend to WD drives at about the same rate. Nobody
is going to think "Ok it's a WD so I will treat it like my
long lost pal instead of like any other HDD".

I could go thru 100 ED drives and very likely not find one that is
bad. Of course, I get to specify the vendor, so I know some gorillas
working for a poor vendor didn't play basketball with the drive.

I think if you took the time to chase down the WD drive failures you
have seen, that in essentially every instance it will be the result of
improper treatment.

I think you have no idea what you're talking about.

I have seen some real junk for half price. Just look at Worst Buy's
weekly fliers.

No you haven't, Best Buy et al has had several drives,
including Western Digital, on sale and with rebates.

Then they need to get a server tower.

Yes, but it has nothing to do with whether their drive is a
WD or not.

That is correct. You cannot put bullets into them and expect them to
work. You have to keep them cool, you have to make sure they do not
get dropped. You can't move them around while spinning (or you
shouldn't), etc.

.... just like with any other drive.

If the data must be current, then you have no other choice but to
implement RAID. But most people do not have that problem so RAID is a
waste of money. They should spend the money they save on a good case
and several good fans.

Can you point out even one case where, given a choice, the
user wouldn't want the "data current"? That is, excepting
the cases where they're reverting to a point just before the
introduction of the problem, with the OS most often.

Again, RAID is usually not as useful for a PC as the
separate backup, but IS still useful for some people and
some uses. If it's not for you then ok, but it would be
silly to contrast your particular uses to someone else's
different use and claim that because your use doesn't
benefit, someone else's wouldn't either.

 

[Bob]

XCOPY meaning the DOS command, right?
Yes.

That's fine for some
but remember that WinXP defaults (and is, on OEM boxes)
running in NTFS, not FAT32.

I am running Win2K with NTFS. I use xcopy all the time.

What OS are you talking about if you're not familiar with
"system restore"?

The jargon "system restore" is not part of the official Windows
lexicon. Therefore I cannot be expected to know what it means.

Look up the phrase "system restore" in the Win2K Help under Search.
You will get a lot of generic topics ranging from "Automatically put
your computer in hibernation mode" to "Remote Installation Services"
to "Windows 2000 Registry".

Again, you are using a common and "official" Windows term,
"rollback" to describe something else. I'm not suggesting
it's a bad term to use based on the context, BUT,
unfortuately when MS usurps some phrase for their own
purposes, it then becomes the common definition since the
majority of PCs run windows.

I thought I defined it clearly in one of my posts.

This discussion has degerated into an argument over semantics. I think
it's time we ended it and moved on.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[Bob]

Neither do I. Put a handfull of drives in a rack, all with
good power and cooling, and sometimes it's the WDs that
fail, not the other brands. If WD were the only brand that
didn't fail, the others would be out of business already.

I said that you have to allow for infant mortality. Now go do your
test after making sure the drives you will use are past infant
mortality.

I have always taken good care of disks and the WD have never crapped
out on me. SMART indicates that WD drives I have been running almost
continuously (24x7x365) for 5 years have so few defects recorded that
the disk looks like new.

We don't have to care, because most people run a
pre-configured OEM box that either had a WD in it or some
other brand. They don't change their system and whichever
drive was installed, could fail. In that most common
scenario, WD drives fail too, at similar rates.

That says nothing about how this OEM box treats drives. If anything,
it likely overheats them. Why do you think I added fans to my case,
which was not a bad case to start with?

It would be ridiculous to claim that the "poor treatment"
doesn't extend to WD drives at about the same rate. Nobody
is going to think "Ok it's a WD so I will treat it like my
long lost pal instead of like any other HDD".

You are making my point for me, namely that most drives are
mistreated. No wonder you find some WD drives among the casualties.

I think you have no idea what you're talking about.

Uh oh - the personal attacks begin.

Time to say goodbye.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[kony]

I am running Win2K with NTFS. I use xcopy all the time.


The jargon "system restore" is not part of the official Windows
lexicon. Therefore I cannot be expected to know what it means.

Lexicon? It's a very common term and anyone dealing with XP
features is usually aware of it.

Look up the phrase "system restore" in the Win2K Help under Search.
You will get a lot of generic topics ranging from "Automatically put
your computer in hibernation mode" to "Remote Installation Services"
to "Windows 2000 Registry".

Ok, sure, it's not a feature of Win2K. Nothing wrong with
using 2K if it fits your needs. Even so, Google for "system
restore".

I thought I defined it clearly in one of my posts.

yes, redefinted it. At that point I'd noted that we had
different definitions of rollback.

This discussion has degerated into an argument over semantics. I think
it's time we ended it and moved on.

I agree.

 

[kony]

I said that you have to allow for infant mortality. Now go do your
test after making sure the drives you will use are past infant
mortality.

Yes, and still the situation is the same. Who puts in a
rack of discs just to take them out again while still in
their infancy?

I have always taken good care of disks and the WD have never crapped
out on me. SMART indicates that WD drives I have been running almost
continuously (24x7x365) for 5 years have so few defects recorded that
the disk looks like new.

That's great, you are lucky. WD drives do fail though,
after infancy, with good power and cooling, good handing.

That says nothing about how this OEM box treats drives. If anything,
it likely overheats them. Why do you think I added fans to my case,
which was not a bad case to start with?

Yes many OEM boxes do run the drives hot, but we can't very
well discuss drive fitness without considering all
scenarios. There is no evidence that WD drives have a lower
failure rate than others.

You are making my point for me, namely that most drives are
mistreated. No wonder you find some WD drives among the casualties.

No my point is that WD drives aren't faring better in either
good or bad conditions, that EITHER WAY they're still
failing at similar rates to other drives.

Uh oh - the personal attacks begin.

Time to say goodbye.

Ok, but consider that you're basically rehashing the same
arguments shown invalid for years. You're considering tiny
sample sizes and trying to draw sweeping conclusions.

Attribute the alternate brands' failure to a cause and show
that the WD alternative in that drive class is immune to
such problem and also doesn't introduce any other weakness
not seen in the other makes.

If WD were half as great (relative to other brands) as you
claim, the rest would be out of business already.

Further, you completely ignore that some drive from a decade
ago has no bearing on failure rates of current, different
models whether they be from same or different maker.

 

[Bob]

Lexicon? It's a very common term and anyone dealing with XP
features is usually aware of it.

I don't deal with XP. I made that abundantly clear earlier.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[Bob]

Yes many OEM boxes do run the drives hot, but we can't very
well discuss drive fitness without considering all scenarios.

That was not the discussion I started. I said that if treated properly
WD drives are extremely reliable. I do not have to consider "all
scenarios" because I ruled out all those in which the WD drives are
not treated properly.

There is no evidence that WD drives have a lower
failure rate than others.

If you do not treat them properly, I can understand that. They are not
designed to be mistreated.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[John McGaw]

Hello, I am building a new sys from newegg products. I would appreciate
any input on the hardware I have selected:
-------------
Antec SONATA II Piano Black Steel ATX Mid Tower Computer Case 450Watt
SmartPower 2.0 ATX snip...
AMD Athlon 64 3200+ Venice 1GHz FSB Socket 939 Processor Model
ADA3200BPBOX - Retail

snip...

The case is excellent from my experience so far. It is indeed cool and
quiet. I've got a fresh system with 4 X 200gB Seagate PATA drives
filling the lower cage and a 160gB SATA drive mounted in an adapter
below the optical drive. To cool the four lower drives I added an Antec
120mm fan in the secondary position to the rear of the cage and run it
at low speed.

I've got a 3500+ Venice which runs quite smoothly. Using the stock OEM
heatsink that comes in the box I can easily run a 10% overclock and
still keep the CPU at 55C and the CPU fan < 1900RPM and essentially
inaudible. This is with the CPU running maxed out with the CPDN project
client and Lame encoder running simultaneously. The only fault I can see
with the AMD so far is that it doesn't handle two heavy loads as
gracefully as a P4 with HT but this machine will be serving basically as
a file server with the CPDN client running in the background so there
should be no power sharing conflicts.

Oh yeah, the cooling snorkel gadget in the Sonata II really IS a drag as
at least one reviewer reported when I need to get into the machine to
change something. That shouldn't be a real problem later when the box is
stable. It does seem to add something to the cooling efficiency in that
it allows the automatic CPU fan to slow just a bit while maintaining 55C.

 

[kony]

I don't deal with XP. I made that abundantly clear earlier.

OK, though even if you don't it's still a common term. I
suppose this is one of the reasons it's good to avoid
off-topic tangents and keep things in a hardware context.

 

[kony]

That was not the discussion I started. I said that if treated properly
WD drives are extremely reliable.

I don't know about the "extremely" part, but yes they are
good drives, I didn't mean to create the impression that I
was suggesting anything else.

I do not have to consider "all
scenarios" because I ruled out all those in which the WD drives are
not treated properly.

Well something more has to be considered because you're
drawing a conclusion that is contrary to far larger users'
(users as-in, any who use, like shops or OEMs) conclusions.

If you do not treat them properly, I can understand that. They are not
designed to be mistreated.

Of course not. There is no evidence that WD drives have a
lower failure rate than others "if you do not treat them
poorly...", either.

 

[Bob]

OK, though even if you don't it's still a common term.

Not common enough for me to know it.

I suppose this is one of the reasons it's good to avoid
off-topic tangents and keep things in a hardware context.

The discussion was about hardware RAID 1 vs. Hardware backups.

You are the one who dragged in obscure jargon.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[Bob]

Well something more has to be considered because you're
drawing a conclusion that is contrary to far larger users'
(users as-in, any who use, like shops or OEMs) conclusions.

That something more is my penchant for babying disks. I mean,
literally treating them with such care nothing will ever happen to
them. Why do you think I have 4 fans in my case. You could hang meat
in there, it's soooo cold.

Of course not. There is no evidence that WD drives have a
lower failure rate than others "if you do not treat them
poorly...", either.

I can believe that. But what I am saying is that if you do treat them
properly they outlast any other drive maker on the average.

They are not the cheapest drive and there's a reason - WD does not
sell seconds under the WD name. Any day of the week you can go to
Worst Buy and get some piece of crap second run disk (albeit with a
brand name) for a lot less. I have been tempted to try it but
fortunately I overcome the momentary lapse into stupidity. I have
always paid up to get WD and I have been rewarded for it.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[kony]

I can believe that. But what I am saying is that if you do treat them
properly they outlast any other drive maker on the average.

You would need a lot more than a subjective experience to
have that be reliable information. Further, experience with
past models of any drive are no indication of modern models'
lifespans. A current Seagate, Maxtor and WD are likely to
have more in common than a WD from 5 years ago and one you
buy today.

They are not the cheapest drive and there's a reason

The price differences are very low and random. I've bought
WD 80GB drives in the past for $20 after rebate. What costs
less than that?

WD does not
sell seconds under the WD name.

Do you have any proof of this, or proof that any other brand
does so?

Any day of the week you can go to
Worst Buy and get some piece of crap second run disk (albeit with a
brand name) for a lot less.

.... or every other day, the WD will cost less. That's
called promos, sales, rebates, etc.

Why are you considering only oddball pricing models like
BestBuy though? How about a competitively priced online
retailer that does mass volume?

You like Directron, how about them?
Check their 'site...

WD 120GB, $70
http://www.directron.com/wd207200.html

Samsung 120GB, $72
http://www.directron.com/samsung1201.html

Seagate 120GB, $78
http://www.directron.com/st3120023a.html

Maxtor 120GB, $74
http://www.directron.com/6y120l0.html


I see a pattern here but not the one you claim.

I have been tempted to try it but
fortunately I overcome the momentary lapse into stupidity. I have
always paid up to get WD and I have been rewarded for it.

I'm not arguing that you should change brands, but simply
buying a WD and waiting for a SMART warning is not a safe
strategy to use. EVERYONE thinks their data is fine till a
drive fails... then they're reminded of why the rest of us
backup data. Data backup is meant to cover those events we
DON'T expect to happen, not the ones we do, thus can take
counter-measures to prevent beforehand.

 

[Bob]

You would need a lot more than a subjective experience to
have that be reliable information.

Actually, my subjective experience is good enough for me - and that is
all that really counts. People can ignore my experience for all I
care.

Further, experience with
past models of any drive are no indication of modern models'
lifespans. A current Seagate, Maxtor and WD are likely to
have more in common than a WD from 5 years ago and one you
buy today.

There's that famous hedge word "likely".

This discussion is going nowhere. Your experience is different from
mine. You sample the experience of the masses and I am limited to my
experience. We are in two completely different worlds.


--

Map of the Vast Right Wing Conspiracy
http://home.houston.rr.com/rkba/vrwc.html

"Beer is proof that God loves us and wants us to be happy."
--Benjamin Franklin

 

[kony]

Actually, my subjective experience is good enough for me - and that is
all that really counts. People can ignore my experience for all I
care.


There's that famous hedge word "likely".

Yes, "likely" becaues drives have the potential to change
from one model or series to the next... another reason why
any past success with your WD can't be applied to current or
future generations of WD drive.

This discussion is going nowhere. Your experience is different from
mine. You sample the experience of the masses and I am limited to my
experience. We are in two completely different worlds.

Well I do agree that mishandling accounts for a large
percentage of failures... though it's not always the user
mishandling, could be a shipper in which case one would have
to buy locally if they have no assurance any particular
shipper will do better than another shipper.