The Blakeyrat Tries To Diagnose This Shitty Work Laptop Thread



  • Windows 7.

    Dell Precision M3800.

    Bluescreens:

    1. When ForteNet VPN was 95% connected

    2. When creating a Git commit in Visual Studio (the Git repo is now corrupted, hah)

    3. When running the Windows "Check for Problems" wizard

    Gonna try and check the Disk Management for SMART problems, without connecting the VPN. This thing has one of those dumb SSD/spinning combo drives in it, and I'm thinking the disk is the most likely culprit.



  • Disk status: Healthy.

    Ok, so far all the bluescreens have happened while Wifi was connected; let's use a USB wifi dongle and see if they happened if I don't use the laptop's built-in network card. If it's the disk, it'll probably crash while installing the dongle's driver...



  • Are those deterministic or not?

    Also, plz give teh bluescreen codez (minidumps would be even cooler, but I assume those might leak some sensitive info, so I dunno). Maybe somebody will barge in with some unsolicited help and save you some work.



  • I am more concerned about who employs you; who gives you that sort of crap and expects you to work with it?

    I only ask so I can put them in cv.robots.txt file :)



  • @Maciejasjmj said:

    Are those deterministic or not?

    Well I can't repro the "creating a commit" one because the Git repo's fucked, and I'm guessing my odds of re-pulling the Git repo without a bluescreen are pretty slim.

    I can try the "check for problems" one again I guess, but first I wanna try disabling the laptop's networking and using this USB dongle.



  • The VPN would be the easiest to reproduce... Did the bluescreens actually say anything sensible? Driver names, perhaps?

    Also, I'm with @loose in that if that's supposed to be your work laptop, tell whoever's in charge of equipment to give you working tools.



  • @loose said:

    I am more concerned about who employs you; who gives you that sort of crap and expects you to work with it?

    What do you like? Dell Precisions are pretty goddamned nice. My last 2 workplaces have given me those shitty Lenovo ThinkPads, ugh.

    Ok it turns out this Linksys dongle doesn't have a driver on Windows Update, so I downloaded it on my personal desktop and moved it over, it's still installing. There it goes.



  • @Maciejasjmj said:

    Also, I'm with @loose in that if that's supposed to be your work laptop, tell whoever's in charge of equipment to give you working tools.

    Well obviously I'm planning to, but I don't want to have to take a PTO day today because I can't get any work done. Jesus, people, do you think I'm an infant or something?

    BTW I just got the bluescreen while restarting after the driver install and it says the component is "memory_management"



  • Ok it looks like about 90% of the time the Memory_Management bluescreen is exactly what it says it is, bad RAM. It is possible to be caused by a driver though.

    The network card Dell put in it disabled in Device Manager, gonna see if this makes a difference in stability but I'm not hopeful. Now almost sure it's actually bad RAM.



  • @blakeyrat said:

    Well obviously I'm planning to, but I don't want to have to take a PTO day today because I can't get any work done.

    Don't you have any spares set up? Huh.

    @blakeyrat said:

    BTW I just got the bluescreen while restarting after the driver install and it says the component is "memory_management"

    Try running Windows' Memory Diagnostics (somewhere in Control Panel, Administrative Tools, under "Diagnose your computer's memory problems" or something like that). Or swap your RAM around, if you have a screwdriver handy.



  • I feel for you, I really do (that's a third leg in my possession - I must stop doing this). But

    @blakeyrat said:

    but I don't want to have to take a PTO day today because I can't get any work done

    Sorta reinforces my feeble attempt at sarcastic humour. Still, it could be worse, you could be working on Discourse.



  • @Maciejasjmj said:

    Don't you have any spares set up?

    Why would I have spare work laptops at my house?

    @Maciejasjmj said:

    Or swap your RAM around, if you have a screwdriver handy.

    1. I'm not even sure how to open one of the Dell Precisions in the new case with the chicklet keyboard

    2. I'm not going to screwdriver around in someone else's PC without at least permission from its owner

    @Maciejasjmj said:

    Try running Windows' Memory Diagnostics (somewhere in Control Panel, Administrative Tools, under "Diagnose your computer's memory problems" or something like that).

    Doing that now, requires a reboot. I've never touched this feature before, is this like a Microsoft version of memtest?

    Before it even starts the test, it says "hardware problems were detected, you will need to contact the computer manufacturer." That's with overall test status 7% complete. So.



  • @blakeyrat said:

    Why would I have spare work laptops at my house?

    A really strange fetish?

    @blakeyrat said:

    I've never touched this feature before, is this like a Microsoft version of memtest?

    Apparently, something like that.

    @blakeyrat said:

    Before it even starts the test, it says "hardware problems were detected, you will need to contact the computer manufacturer." That's with overall test status 7% complete. So.

    Well poop, enjoy your day off I guess.



  • What was the last thing you done (physically) to the laptop?



  • Yeah pretty much. Hopefully the IT guys at work can bring it back to life with a RAM transplant, if I have to reinstall everything that's like an entire day lost.



  • @loose said:

    What was the last thing you done (physically) to the laptop?

    Put it in a backpack and brought it home?

    I wonder if it has something to do with the heat. It was probably 90 or 100 in that black backpack Friday afternoon.

    The memtest utility causes some kind of bitlocker error when it tries to reboot, but rebooting manually seems to work ok.


  • sockdevs

    @blakeyrat said:

    Ok it looks like about 90% of the time the Memory_Management bluescreen is exactly what it says it is, bad RAM.

    i doubt this is the cause but have you checked that the amount of RAM in the computer matches what windows sees?

    I ask because i was issued a laptop that used two 8GB sticks, on a motherboard that only supported up to 4GB per DIMM. It would regularly BSOD on resume from hibernate because resume from hibernate would see the memory mismatch and freak out about it... It's possible you've got something like that going on here?



  • Why would Dell ship that?

    Anyway this is unrelated to hibernate, and I highly doubt Dell put in useless RAM. Dell might not be the best PC maker, but they're not that dumb.



  • Probably another attempt to drive down the value of your Amazon stock.



  • If it was a Pc I would suggest a general reseat. But like you said. Tamperproof Laptop Case, and not yours etc. Does it not have a little "slide and click" type thingy access covers to make memory /disk upgradescheapexpensive?


  • sockdevs

    @blakeyrat said:

    Why would Dell ship that?

    they didn't, for one this was an HP and for two it was a reissue and all Helpdesk had to replace the bad ram were 8GB sticks.

    IT gets the crap laptops here because we only need them for meetings, our actual dev work is done on our desktops. the rest of the company is only issued the laptops so they're higher on the totem pole for that hardware.



  • Ok I rebooted and was looking forward to see the memory test results but the screen has frozen, haha.

    I can see the mouse move (which I believe is done by hardware on most machines), and the USB network card is networking away so the computer's running at some level, but nothing's being drawn to the screen.

    Oop, now it's locked-up solid. Oh well, I give up.

    I'll reboot long enough to get the networking hardware turned back on so the IT guys don't call me an idiot when I say it's not working, send them a heads-up, then I guess PTO day. My boss is gonna be pissed, this is like the 5th day he's lost to bad hardware in 2 weeks.



  • Ok we'll see how my boss replies. I told him I could drive the laptop down to Seattle around noonish if he thinks it's a huge deal, but I'd actually rather not if I can avoid it.

    Also gave a heads-up to IT that it's coming. So. Whee. Thank God for webmail, I guess.

    BTW it's this model:

    For a 15" laptop, it really is quite slick. When it works.



  • In my experience, frequent bluescreens = bad RAM in 90+% of cases.

    Now to read the thread.



  • @blakeyrat said:

    Lenovo ThinkPads

    They're convinced that Lenevo offers the most business productivity per dollar I suppose.
    It's taking over everywhere.



  • I've had a couple Acer machines and both have been rock-solid, but for some reason Acer has nearly zero penetration into the business market.



  • Ok it finished a full chkdsk, fixed a bunch of problems, dare I see how far I can get? Maybe I can at least read that JIRA ticket and work on it offline?

    EDIT: Aaaaa! I got like 90% through explaining why the JIRA ticket is user-error before the thing bluescreened. Ok now I give up for realsie.

    Thanks all.

    Maybe I can reply to JIRA by email.



  • @blakeyrat said:

    Now almost sure it's actually bad RAM.

    Bingo.

    @blakeyrat said:

    I'm not going to screwdriver around in someone else's PC without at least permission from its owner

    Wuss. If you're even slightly careful they'd never know you'd done it.

    @blakeyrat said:

    I wonder if it has something to do with the heat. It was probably 90 or 100 in that black backpack Friday afternoon.

    If it was the bottom of your laptop that copped all the heat in that bag, then fixing it might be as easy as re-seating the RAM cards. With any degree of attention paid at all to avoiding ESD, you could not conceivably make it worse by trying that.

    Last laptop I had with a RAM fault actually turned out not to be the RAM, but the socket. A kid had sat on the thing, for which the customer had had the screen replaced by somebody else, but it was still flaky; I popped out the RAM cards to check the edge connectors for thumbprints, but noticed one connector finger in one socket looked like it was sitting lower than the others. Bent over the very tip of a ball-headed pin and levered it approximately back into place. After a trial refit and removal of the RAM card it looked like it was still in the right place, and the machine then passed 24 hours of Memtest86+ (it had initially failed consistently in the first pass).



  • @blakeyrat said:

    it finished a full chkdsk, fixed a bunch of problems

    Running chkdsk in a machine with bad RAM is actually more likely to cause problems than fix them.



  • Yeah I thought about that about halfway through, but since Windows decided to schedule it and since by the time I thought about that it wasn't really cancel-able, ... eh.

    If the filesystem is fucked now, it's Windows' own fault for scheduling a ChkDsk after its memtest already knew about RAM errors.



  • @blakeyrat said:

    I've had a couple Acer machines and both have been rock-solid

    My school has been buying Acer consistently since 2005, purely because one of their local resellers is just such a pleasure to deal with (Learning With Technologies, for other Australian readers - I have nothing but praise for this outfit).

    I've seen a pleasingly low failure rate on their desktop workstations: three or four mobos with capacitor issues, a couple of dead power supplies, two or three bad RAM sticks in over 400 machines that have been through the school since then (we run them for about five years), all faults promptly repaired under warranty. I've also had an Acer server whose twin PSUs, disk controller and Ethernet card all died at about the same time, possibly related to something horrid that our electricity supplier was doing to the pole transformer up the road; AcerCare tech support was first-rate, didn't treat me like an idiot, and once we figured out that the fastest way to get a machine 320km from the capital city repaired was going to be for them to ship me parts, did so quickly and willingly.

    Haven't had many of their laptops - maybe fifty? One DOA, one bad battery, all dealt with promptly under warranty.

    Acer's prices are pretty good for what their machines are, and I've found both their local supplier and their local tech support folks much better to deal with than either Dell's or HP's.



  • Sounds like classic MS Unit Integration: Right not knowing what the left hand is doing


  • mod

    @blakeyrat said:

    Maybe I can reply to JIRA by email.

    That's a feature JIRA has, but your admins have to enable it, so it's a toss-up whether you can or not.


  • Discourse touched me in a no-no place

    @blakeyrat said:

    I'm not even sure how to open one of the Dell Precisions in the new case with the chicklet keyboard

    Dells usually have a panel on the bottom you can remove to get at the RAM, but with your point #2, it's probably not worth doing. (Except if there were two DIMMs and taking one out made the problem go away, you know what's wrong and what the fix is, and IT should appreciate that.)



  • There's an old model of Dell laptop where the ram sticks are located in two different places, as I found out when I was upgrading it years ago. One stick of ram is behind a panel on the back, like you'd expect. The other one is under the keyboard, meaning that you pretty much have to take apart the whole thing to get at it.



  • @Yamikuronue said:

    That's a feature ert product name here has, but your admins have to enable it, so it's a toss-up whether you can or not.

    Wow, that sentence works for several products!


  • Discourse touched me in a no-no place

    @hungrier said:

    There's an old model of Dell laptop where the ram sticks are located in two different plac

    Dell has a history of weird shit.

    All the machines we have here at work are Dells and none of them are weird--I think they outgrew oddball designs at some point.



  • I would hope so.

    Apparently the ridiculous "under the keyboard" design is/was more common than I thought: Here is a guide with lots of pictures on a different model of Dell than I had, but basically the same.



  • I hope you're not still posting here assuming I'm going to open up this laptop. Because I ain't.


  • Discourse touched me in a no-no place

    My first post was having only read up to that point.

    I don't blame you for not wanting to open it up, btw. I probably would've, myself, but only if I'd been at the company for long enough to have a feel as to whether or not it'd be OK.



  • I just wanted to share something ridiculous and slightly relevant.



  • Took laptop in to IT, they ran a Memtest and determined that only the second stick of RAM they put in the machine a few months ago as an upgrade was faulty. So he removed that, I'm back to 8 GB but it... seems? to be working?

    He's got a computer ready to image if it bluescreens again, but the imaging takes an hour and change and I'd have to reinstall everything, so. We'll see.

    Now the problem is my Git repo is corrupted, so I have to repull that. Ugh.

    EDIT: no wait, both VS and Skype just crashed out of nowhere for no reason. This laptop's still funky. Ugh.



  • Haw haw.



  • I'm just seeing if it can run reliably enough to clone a 1.1 GB git repo. If it can do this, it can do anything.



  • The Precisions my company's IT gave us are a crashy downgrade from the previous set. None of the employees with the last set are willing to upgrade to the ones we got, since they have SSDs, whereas we have hybrid drives like you. Mine sometimes bluescreens (twice ever, or I'd probably have a new one by now) when I log into Windows, and has an unstable fan that becomes loud at random times.

    But I don't honestly trust this place's IT to fix it or give me anything that works better.



  • OK!

    He re-imaged the same laptop and returned it to me. If it can survive a VS2013 and a SQL Server install, then I guess it's probably reliable enough.



  • Running pretty solid now.

    I guess the memory corruption led to disk corruption. Hopefully not due to the chkdsk that ran under my watch, but oh well.

    VS installed, SQL Server is next.



  • @blakeyrat said:

    SQL Server is next.

    We'll see you in a few hours, then! <worst installer ever...


  • Discourse touched me in a no-no place

    @blakeyrat said:

    EDIT: no wait, both VS and Skype just crashed out of nowhere for no reason. This laptop's still funky. Ugh.

    I wonder if it's the motherboard (or whatever the equivalent is for laptops) that's carked it. I've seen that on a workstation here at work (it was Solaris system about 15 or so years ago) and we had the maintenance engineer out about 5 times trying to figure out what was wrong. It would work every time he changed something (memory, disk, processor, etc.) until a few minutes after he left the building. :smile:

    When we finally got the motherboard replaced, that fixed it. Fortunately, all my colleague's programs and data was stored on networked filesystems so he didn't lose anything. Reimaging was very quick. (I wouldn't advise keeping applications on networked filesystems these days — if the network goes down, you end up really hosed — but it made some things quite convenient.)



  • Just when I finally got my dev environment all installed and set up-- BOOM another bluescreen.

    Now I have to do it all over again. At least this is a new laptop, straight out of the box. Because I'm sick of it.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.