Jump to content

hogfather

Member
  • Posts

    143
  • Joined

  • Last visited

Awards

This user doesn't have any awards

Contact Methods

Profile Information

  • Gender
    Male
  • Location
    Prague, CZ
  • Interests
    Biking, IT
  • Biography
    I enjoy biking and IT.
  • Occupation
    IT manager

System

  • CPU
    4930K @ 4.5ghz
  • Motherboard
    Asus Rampage Black
  • RAM
    Corsair 64GB
  • GPU
    EVGA Hybrid 980ti
  • Case
    Corsair 750D
  • Storage
    SSD: 1x Kingston 256GB, 1x Crucial 256G, 1x Crucial 128GB, HD: 2x Seagate 2TB
  • PSU
    Corsair 850HXi
  • Display(s)
    Philips 27" 1440p
  • Cooling
    Corsair H105
  • Operating System
    w10

Recent Profile Visitors

1,120 profile views
  1. Thanks for the reply @Jarsky. Good point on the firmware. I tried ' sudo isdct show -intelssd' and it produces : "FirmwareUpdateAvailable : No known update for SSD. If an update is expected, please contact your SSD Vendor representative about firmware update for this drive." Which is actually correct, because as far as I know my drives are the 3605 versions, which were for Oracle (from memory). I might try and find some Oracle tool to check the firmware with. The calculation your provided make total sense, but I really don't trust the number from smartctl. Anyway, thanks a bunch for taking a look.
  2. Hello I was hoping someone with experience with these drives, or similar, can shed some light on what I believe are incorrect smartctl numbers. I bought two such drives from ebay for like 100 EUR each and they live a pretty hard life in my home lab where they are punished with mainly write intensive workloads. The workloads are mainly my dev Kubernets setup which runs multiple Apache Pulsar clusters. From the stats I noticed that the "Data Units Read" and "Data Units Written" often reset back to 0 when they go past about 2.2 - 2.5TB. I would imagine that these should be an aggregate of all data written to the drives. One drives shows it's only been used 2%, but it's been on for 6.5 years. I find that hard to believe and I actually have no idea how much data has been written to this particular drive. The "Power On Hours" and "Power Cycles" seem to update correctly. I read some posts on other forums where people we saying that these stats can be off. Is that true? At the end of the day, I don't mind because none of my drives store important data for long. If anything the data is stored for a few hours before getting wiped and rewritten. On the topic of writing, I'm certain that I have written more than 500 TB to the drive below which should amount to 5.7% (1.6TB @ 3 DWPD = 8760TB written), yet the "Percentage Used" hasn't moved. $ sudo smartctl -a /dev/nvme0 smartctl 7.2 2020-12-30 r5155 [x86_64-linux-5.15.0-88-generic] (local build) === START OF INFORMATION SECTION === Model Number: INTEL SSDPEDME016T4S Firmware Version: 8DV1RA13 PCI Vendor ID: 0x8086 PCI Vendor Subsystem ID: 0x108e === START OF SMART DATA SECTION === SMART overall-health self-assessment test result: PASSED SMART/Health Information (NVMe Log 0x02) Critical Warning: 0x00 Temperature: 34 Celsius Available Spare: 100% Available Spare Threshold: 10% Percentage Used: 2% Data Units Read: 4,247,474 [2.17 TB] Data Units Written: 3,160,725 [1.61 TB] Host Read Commands: 144,368,876,738 Host Write Commands: 21,241,519,364 Controller Busy Time: 1,303 Power Cycles: 62 Power On Hours: 57,074 Unsafe Shutdowns: 3 Media and Data Integrity Errors: 0 Error Information Log Entries: 0 Error Information (NVMe Log 0x01, 16 of 64 entries) No Errors Logged Any ideas or thoughts? Cheers!
  3. I would avoid using either a consumer CPU or storage for this kind of build. I'd also pass on the top of the line CPU for any generation unless is has significantly come down in price. My justification for being against a consumer CPU is that you will be restricted by the number of PCIe lanes and the amount of RAM you can install. On top of that, a 5950X is still recent so it still has some time before its price drops a lot. A consumer CPU won't have enough lanes to handle a graphics card (or multiple of), PCIe/U.2/U.3 drives, presumably a faster NIC (a 10gb or a 25Gb even) and any HBA(s) for larger storage arrays. The reasoning based on my experience: I had a 4930K which started to run really slow so it was time to replace it. I decided against selling my Asus Rampage 4 BE and instead I spent 70 EUR on a E5-2670 v2 (https://www.cpu-world.com/CPUs/Xeon/Intel-Xeon E5-2670 v2.html) back at the start of 2021. Back then a top of the line for that generation was a E5-2697 v2 (https://www.cpu-world.com/CPUs/Xeon/Intel-Xeon E5-2697 v2.html) which was going for around 150 EUR. The E5-2670 v2 did well enough for me to host an mdadm raid of 4 HDDs (using onboard SATA only), a small MySQL DB for my weather station, pihole, a number of Docker containers for various bits and pieces and 5 qemu virtual machines which made up a test Kubernets cluster. At the beginning of this year I bought the E5-2697 v2 for 35 EUR (the E5-2670 v2 can be had for under 20 EUR now) and that is more than enough for all the dev/ops things I do. In terms of storage, the 5 qemu virtual machines lived on one Crucial SSD and every 24 hours of run-time they would eat up 1% of the SSD life. After less that 6 months the SSD's life was all but gone. So I replaced it with a Intel S4610 and never looked back. One of them set me back 100 EUR and I got two. One for the VMs and one for the main OS (I know, it's excessive for the OS, but I had a crappy Kingston SSD which also needed replacing). For block storage I picked up 2x Intel P3605 PCIe NVMEs as the stuff I work on is storage bound. Both set me back a total of 200 EUR. Both drives have over 50K running hours, but still have more than 80% life left based on the number of writes they have endured. This is ideal for my setup as these drives get wiped multiple times per day and never store any important data. On any given day I'll easily write more than a few TBs of data to each drive during my tests. Over 1 particularly intensive weekend of testing each drive wrote and read over 200TB. Just keep in mind that cooling server components is a thing and can be tricky in a normal case. I use an old Corsair 750D and I had to Frankenstein a bracket to point a fan over the NVME drives to reduce their temps by ~12 degrees (C) As my motherboard is getting on in age now, I'm already looking for a future replacement and as of now that is looking like a jump to LGA3647. You can pick up a https://www.cpu-world.com/CPUs/Xeon/Intel-Xeon 6130.html for around 100 EUR give you 16c/32t while you wait for the other monsters in this generation to come down in price. Dual CPU boards are also an option on this socket and one I'm willing to consider when the times comes. All in all, it all comes down to your exact use case and there many ways to skin this particular cat.
  4. That did the trick! thank you! Now you can go back to trying to take over middle earth.
  5. First of all I'm no mathematician. Second of all I have the answers to my problem. What I'm after is a formula, not because I need it, rather because I'm infuriated by not being able to come up with one myself. The problem is this: Given any combination of letters (as in a word), come up with all the possible permutations and combinations. That's fairy straight forward to solve using a factorial (think that's what's called). for example the word 'hello' gives you 120 ways of arranging all 5 letters into unique positions (5! = 120 and ignoring the double 'l'). What I'm after though, is how many unique combinations can you have including combination that are made up with less letters. For example the word 'fox' gets you 15 such possibilities. I won't list all of them, but basically you combine all the letters in which ever way, so even the letter 'x' on its own is a viable result. I know that for 4 letters you have 64 such combinations, for 5 170 and for 6 you end up with 1050. I don't need to go to infinity, i just need to go as far as the longest word in the English language (not sure which particular variant just yet, but probably British English). The reason I'm doing this: Simple - I was reading Dan Browns The Davinici Code and there the main female's characters grandfather was dishing our word puzzles and of of them was how many words can you form from all the letters in a given word. So i thought I'd code up something like that, which wasn't difficult, but now i'm struggling with a formula. The code I have does find all the possible word combinations for a given word, but I'm obsessed with knowing a formula. I've look at formulas for permutations and combinations but all I've found is suggestions that always give you the same number of letters. I think my case is a bit more complex, but since I'm not a mathematician, I can't get my brain around a formula for it. Any help is appreciated!
  6. Been using pi-hole for more than a year now and we don't have any ads at home on any of our 10 devices. I went out bought a cheap Intel Pentium G4560 and stuck it in a case under a shelf. It runs Ubuntu and is easy to set up. On average around 30% of traffic gets blocked but of course your millage will vary with your specific blacklists. All in all, I'm super happy with pi-hole.
  7. I can't get my head around the blame of an ad-blocker for the floatplane issue. Not even sure that issue was high-lighted on this show on the one from the previous week. Either way, I cringe when I see the screen share covered with BS ads. I recall one episode where the team was talking about how they don't use an adblocker at LLT. Which 'internet' are you using? Cause the one I'm on is full of crap that my eye balls do not want to be concerned with. In fact, since setting up Pi-Hole, the 'internet' I use has been significantly more bearable. Die ads die
  8. man, Linus, you need to get some sleep! get rid of them rabbit eyes
  9. I'm using digital TV from my ISP. Their website is not that up to date, which is kind of limiting.
  10. I've been having an issue where my install of Ubuntu 16.04 LTS encounters hard system freezes when playing flash content. By hard system freezes I mean that the system locks up, replays the last second of sound and has to be restarted using the reset switch. Some points: I have not been able to get anything useful from the system logs when examining them after a crash. The crashes are random. Sometimes they happen twice per day, while other times once per month. I'm using Chrome and have the Hardware Acceleration turned off. My GPU is a 980Ti using NVIDIA binary driver - version 357.66 from nividia-375 (proprietary). I gone through a few different revisions (open source: 384.47, 370.28, 378.13, 381.22 and the Nouveau drivers) to no avail. Chrome version is: Version 59.0.3071.115 (Official Build) (64-bit) OS (lsb_release -a) : Distributor ID: Ubuntu Description: Ubuntu 16.04.2 LTS Release: 16.04 Codename: xenial I've also seen the freeze occur when watching flash content in a VM running Windows 7, which really surprised me. Is anyone able to shed some light on this?
  11. I would attribute the rise of Win XP growth to the Royal Navy. Isn't their new tub running Win XP?
  12. Holla People I have an odd requirement. I would like to make my EVGA Hybrid 980ti quieter because the blower fan is too loud for my liking. I can see two options: take the shroud apart and put a resistor on its' cable to reduce the rpm, or replace the shroud with a 2/3 fan cool unit. I don't want to sell the card and doing a swap is a bit of a pain in the ass. Has anyone tried doing this?
  13. That's quite a lot TB-age lost due to Windows and Unraid/ZFS. Is that justifiable? Can the loss be, well, lessened somehow? BTW, AWS at these amounts of data is not viable at all.
  14. could be that the paste just dried up. change it. if it doesn't help then it;s a faulty sensor.
  15. swap the cards around and see how they go
×