SSD Drives in Storage Somewhere?



  • Anyone know if there are any SSD drives in storage? Would be nice to swap out the drives in the floor computers (Dell Optiplex 9010's) for faster performance if the equipment is available.


  • Range Volunteers

    @jdez I highly suggest you dont do this.



  • Don't have my heart set on it, but why would that be a bad thing?


  • Core Volunteers

    I'm just learning more about the differences between SSD and HDD, so correct me if I'm wrong. SSDs have flash memory and typically degrade faster than HDDs. It's great for personal use, but we re-image these computers regularly and users work on them all day. Mechanical drives would be better in the long-run. Am I on the right track?



  • @Kaela yes... We've burned through SSDs in the past which is probably why @ccc recommended not doing this.



  • @josh @Kaela Thanks guys...I was aware they didn't last as long...just wasn't aware the issue was bad enough to avoid them here - never compared in real-life and my understanding was they marginally worse.


  • Range Volunteers

    Not only that, its a waste. Its only faster performance if you are shutting the computers off and booting them all the time. Once a computer is booted, the OS in living in memory. Unless you are doing a ton of reads/writes its not going to be noticeable. And even if you did put SSDs in there, your bottle neck is really the CPUs.

    Trust me, SSDs are really only needed for gaming.



  • @ccc Ok...good to know. They just seem to hang after not being used for a while, and when people come in they're jumping from computer to computer finding the one that reacts first.


  • Range Volunteers

    @jdez that isnt a harddrive issue usually


  • Core Volunteers

    @jdez Hmmm. Well my technical knowledge is barely enough to have my first IT job, but we have those workstations running all through the day and I'm guessing they're often lagging because of cache and/or a bunch of background processes still running that don't need to be. The next time someone complains about a workstation lagging, I'd just restart the computer to flush all the junk. I don't know what your skill level is with Linux, but you can also use this as an opportunity to get more comfortable with commands to check processes if you're not already proficient: top, ps -aux, and kill. Then use the pipe | command to narrow your results. 🙂



  • ive been trying to locate the problems of this exact issue for a little over a week and resets only seem to work about half the time, wondering if this is more a hardware issue with all the old pieces of hardware. Anyone know if there is a command to check the cpu temps or is that not supported on the motherboards in the stations.



  • Fun issue to try and track down. My linux isn't the worst, and it's probably a lot further from being the best 🙂 I have a feeling that it's a combination of hardware and initial startup/access of vm components coming from our servers.

    I know Kali 19.1 has some sort of hardware temp tracking application since a gui popup keeps telling me the drive is overheating on my kali laptop, but it's been overheating for a couple years due to the design and video card. I just now did an apt install of hddtemp on my ubuntu laptop and seems to work as long as you specify the drive. I'm betting there's a command to specify/list all drives at once somewhere in the manual.

    [email protected]:~$ sudo hddtemp /dev/sda7
    /dev/sda7: ST9500325AS: 35°C



  • @Virella @Kaela This link includes some scripts for shutting down linux boxes if temps exceed a certain level...couldn't help but think it could be an easy way to screw with someone by adjusting the temp level input.





  • Hey gang. Try this if you're thinking something's wrong harware wise.

    StressLinux



  • ill try that on a couple ive noticed that seem to not change much on restarts, not sure if the hd's are new enough for the hddtemp sensor to be installed in them,. i may also try
    https://askubuntu.com/questions/15832/how-do-i-get-the-cpu-temperature
    incase its a cpu instead of a hard drive issue


  • Core Volunteers

    That's the thing, too. There are so many variables. We have all these workstations running throughout the day and they contain older components. We're using Virtualization and a lot of other technologies. It's hard to pinpoint problems and note one thing as the cause.


  • Range Volunteers

    @Kaela reinstall the OS as well. Often Kali is just plain crap. You guys might just want to switch to Xubuntu


  • Core Volunteers

    @Tails Stress test is a good idea. We have one workstation in particular that definitely has issues. Thank you!



  • @Kaela
    You're welcome. I hope it helps the group 😁



  • done some checking looks like system 19 may have a cpu problem, idles at 40%ish system on one of the cores with nothing running, using the same program i got almost nothing on the other 2 systems i tried it on, the cpu is also running about 20 degrees hotter. i used lm-sensors to get the temps and nmon to check out the cpu usage, ran a bunch of other tests, only thing out of the ordinary was cpu. I shut it down for now want to try it out later see if it changes after it cools off.



  • for anyone wanting to look at the programs i used nmon and the lm-sensors from the earlier post i made, doing some comparison the problem one really stood out



  • I know I'm late to the party, but I have to disagree with much of what's been said here.

    First of all, disk is, in fact, the main bottleneck in AZ03, especially on Windows. Believe me, I've checked.

    Secondly, SSD technology has gotten far better in the past few years - for example the Samsung 860 Evo (500GB) has a TBW of 2400. The largest Fog image we have is under 15GB. Supposing we reimaged every day, and every day 15GB of other data was written to these SSDs, even if "every day" includes weekends, it would take almost 220 YEARS before we hit that.



  • @virtualdxs Assuming you're right about the drives, what's the oldest SSDs that would give us a reasonable TBW? All the range stuff is donated and we'd likely be salvaging from older equipment.

    Thinking logistically, it only takes maybe 3 minutes to swap out disks, and old ones can be stored as backups.



  • To add to that, two CIS labs hosting somewhere between 6 and 10 classes, all using 1-3 VMs that get redone once a semester, plus the hosts get reimaged about once a year, have been running on mirrored SSDs for the past 5 years. We have yet to see even one of them die,



  • @jdez I'm not sure, but we may have a bunch of the ones used in the CIS classes available; working on figuring that out.


Log in to reply