Post your hard drive Power On hours!

[SamirD]

I have discovered a terrible bug in the 1BD142 series units, when they reach 80,000h the counter is reset: '(.
Mine already goes for 77,000h but it has not yet been reset, but my uncle has one like it, and it exceeded 80,000h and it was reset, now it marks 700h.

That's one heck of a bug!

I wonder what it is about 80k poh because at one point there was an HGST firmware bug in some of their enterprise ssds that would brick them at at poh (if I'm remembering correctly, which I'm probably not).

 

[LOCO LAPTOP]

Daayyyummm...that system and drive are awesome for just working that long. I wouldn't have retired it and just left it as a failover for the VM version. Or at least that setup needs a new job--it's clearly still good for one.

We still have well over 150 of those and 5015's (P4 socket 775 versions). Most of them still work, some have dead power supplies, 2-3 have dead boards.

Most of them are retired, I think 20 to 30 of them are still running. We are moving data centers in 2 years, by the end of this year they will all be offline.

Old picture in 2012 shortly after I got hired, that shows half of them lol.

 

[SamirD]

Unbelievable great how those are still running this many years later. Pop in a sound card and use it as a retro gamer.

What surprises me from a business standpoint that they weren't replaced sooner since more efficient servers would have been able to handle the same workload with less hardware aka less power/cooling/etc.

 

[LOCO LAPTOP]

Unbelievable great how those are still running this many years later. Pop in a sound card and use it as a retro gamer.

What surprises me from a business standpoint that they weren't replaced sooner since more efficient servers would have been able to handle the same workload with less hardware aka less power/cooling/etc.

Listen, it took me 2 years of bitching about the CyberPower 2200VA (4 per rack) randomly crapping out taking 1/4 of the rack down every Wednesday's (Generator switch over every week) to get a APC Symmetra PX 80kW installed. It took another year to FINIALLY start getting newer servers with real cabinets and real network switches (Not Netgear trash that couldn't handle a internal file transfer that was maxing out the 1Gig link which cause it to have massive packet loss), which were HP DL360 G6 off lease servers on eBay.

Once those were live and I started to tell them we can offer VPS servers for extra income and I had taken a whole rack of servers into 10 servers (with room to grow), with less power, faster response time, and noise they started to noticed, "Wow, we were really doing it wrong." Almost none of our backend stuff is on hardware that old now. Most of what is left running is leased dedicated servers, very few server 2003 (Yes, still. :/ ), the reset is Linux.

I've been there for almost 10 years. This year, I've been told we are actually getting brand new servers for the new data center. It's not they couldn't afford it, it's they didn't want to.

 

[pitingres]

Listen, it took me 2 years of bitching ...

Many businesses operate using the theory "If it ain't broke while doing the CEO's key Monday report two weeks running, don't fix it." If there isn't someone watching all of the costs, direct and indirect, a lot of bad decisions will be made.

 

[LOCO LAPTOP]

Many businesses operate using the theory "If it ain't broke while doing the CEO's key Monday report two weeks running, don't fix it." If there isn't someone watching all of the costs, direct and indirect, a lot of bad decisions will be made.

I wish they would do that with our apps. They finally get it right then a month later, "here brand new from the ground up version" that breaks everything!

But yeah, you're right. It's just when it goes down, it's our fault. lol

 

[SamirD]

Listen, it took me 2 years of bitching about the CyberPower 2200VA (4 per rack) randomly crapping out taking 1/4 of the rack down every Wednesday's (Generator switch over every week) to get a APC Symmetra PX 80kW installed. It took another year to FINIALLY start getting newer servers with real cabinets and real network switches (Not Netgear trash that couldn't handle a internal file transfer that was maxing out the 1Gig link which cause it to have massive packet loss), which were HP DL360 G6 off lease servers on eBay.

Once those were live and I started to tell them we can offer VPS servers for extra income and I had taken a whole rack of servers into 10 servers (with room to grow), with less power, faster response time, and noise they started to noticed, "Wow, we were really doing it wrong." Almost none of our backend stuff is on hardware that old now. Most of what is left running is leased dedicated servers, very few server 2003 (Yes, still. :/ ), the reset is Linux.

I've been there for almost 10 years. This year, I've been told we are actually getting brand new servers for the new data center. It's not they couldn't afford it, it's they didn't want to.

Wow, I'm stunned that they didn't run their own numbers and look at what the costs were. But if they were getting a deal on the used servers that counts too so maybe it was the same profit either way so why change it, which makes sense to me as that's what I do in business as well.

 

[pitingres]

Wow, I'm stunned that they didn't run their own numbers and look at what the costs were. But if they were getting a deal on the used servers that counts too so maybe it was the same profit either way so why change it, which makes sense to me as that's what I do in business as well.

LOL! IT guy: "Costs are for the accounting people." Accounting: "Why tf do you think I have any idea what goes on in IT? I just pay the bills."

It takes someone who knows not only what the costs are, but which costs are tied to what potential capital expenditures. it takes someone who sees more of the big picture and a distressing number of businesses don't see the value in that. (Even the big ones, because this sort of costing tends to be departmental. Then you have not only small-group-think, but keep-heads-down think.)

 

[SamirD]

LOL! IT guy: "Costs are for the accounting people." Accounting: "Why tf do you think I have any idea what goes on in IT? I just pay the bills."

It takes someone who knows not only what the costs are, but which costs are tied to what potential capital expenditures. it takes someone who sees more of the big picture and a distressing number of businesses don't see the value in that. (Even the big ones, because this sort of costing tends to be departmental. Then you have not only small-group-think, but keep-heads-down think.)

lol, and then you have china come in with all operations under 'the party' and undercut and drive these companies out of business...

 

[toast0]

 smartctl -a /dev/sda  | grep -E '(Device.Model|Power_On)'
Device Model:     WDC WD20EARX-00PASB0
  9 Power_On_Hours          0x0032   001   001   000    Old_age   Always       -       83384

83384 hours = 3474 days = 9 and a half years

I guess I should probably finish retiring this system; it was my home server, but I've moved most of the stuff onto newer systems, it's still the mythtv backend and one of the frontends though.

 

[TimTheSettler]

That's one heck of a bug!

I wonder what it is about 80k poh because at one point there was an HGST firmware bug in some of their enterprise ssds that would brick them at at poh (if I'm remembering correctly, which I'm probably not).

Make a note of the serial number of the drive and take a screenshot. That way, when it resets you have proof.

 

[SamirD]

Make a note of the serial number of the drive and take a screenshot. That way, when it resets you have proof.

Too many people have tools to reset these today.

 

[kevinwoxd]

That's one heck of a bug!

I wonder what it is about 80k poh because at one point there was an HGST firmware bug in some of their enterprise ssds that would brick them at at poh (if I'm remembering correctly, which I'm probably not).

Yes, unfortunately, I hope mine doesn't happen to him, but I don't think he's saved from the error .

 

[kevinwoxd]

Yes, unfortunately, I hope mine doesn't happen to him, but I don't think he's saved from the error .

@SamirD The reset just happened, I am very sad :'(, although in the WORST value it continues to subtract, so it is an indicator that the HDD continues to accumulate hours, but the hour counter was reset. The reset was done at 78183h.

But it is not completely lost, in flight hours the head stores another counter, but it can be viewed with gsmartcontrol.

 

[lilfiend]

Not all of my drives but the longest running of them. The ones with 0's for load/power cycles are SSDs.

 

[D-EJ915]

My file server drives (dt01aca3) are all 61446 hours, no issues yet.

 

[kevinwoxd]

Still alive! It refuses to die, let's see if it can get past 100,000h, and if it does, I'll keep using it until I say enough.

 

[TheSlySyl]

View attachment 515334
Still alive! It refuses to die, let's see if it can get past 100,000h, and if it does, I'll keep using it until I say enough.

What software is this? I really like how this table shows reasons for the failures.

 

[kevinwoxd]

What software is this? I really like how this table shows reasons for the failures.

It's called Gsmartcontrol, It is without a doubt the best open source SMART reporting program, you can also read the errors that a HDD stored in the error tab.

 

[MentalElement]

Power on Hours : 1032 days .. since end January 2014

 

[Zarathustra[H]]

The drives in my ZFS pool in my backup server should be hitting 70k hours any day now.

I should take a screen shot when they do.

 

[Sprung]

My last HDD died sadly, but my current SSD while not as hard worked as most of the ones here is holding out well.