Testing disks

Playing with the HBA:

480MiB/s but I ran out of disk bays.

dd_rescue / Linux seems to like suck memory with this test but at least that 64GB goes into some good use finally.

But performance scales well with now 6 more disks under test:

557MiB/s reading speed.

Linux and SAS3081E-R


These are good devices. With these I can hot swap all of my disks and do all maintenance without any hardware RAID configuration, which with SmartArray required booting multiple times! HP might provide some configuration tool for their SmartArray but with this I need none.

Once I get the other controller I think I have to get couple more disks to store the data temporarily to and then move all the drives in each mirror to different controller. That way I will have no problems even if a controller dies. The operating system disks will have to use hardware RAID 1 because any other solution on Linux isn’t going to be without problems.


New disk bays online + future changes

The new 8 disk bays are now online with LSI 1068E and it seems to just pass the disks right to the operating system so I might have to get another one to replace the HP SmartArray P410 which requires me to use RAID0 which in turn causes problems because now there is that extra HW RAID getting involved with everything such as disk failures and so on.

I just hooked some old 32GB disk to try it out and it seems to work perfectly:

dd_rescue: (info): read /dev/sdg (35566480.0kiB): EOF
dd_rescue: (info): Summary for /dev/sdg -> /dev/null
dd_rescue: (info): ipos:  35566480.0k, opos:  35566480.0k, xferd:  35566480.0k
                   errs:      0, errxfer:         0.0k, succxfer:  35566480.0k
             +curr.rate:    39202kB/s, avg.rate:    52107kB/s, avg.load:  5.6%
             >-----------------------------------------< 100%  TOT:  0:11:23

I think I will just order the card right away. 14.39 € so practically free. After which the server is just about as stacked as possible. Probably still worth less than the price of electricity it will save in 18 months.

Seems suspicious

I don’t know why this seems suspicious, or does the person selling this not understand how markets work. Because the price of silver has come down for the past half a year, and then the person suddenly decides that it is time to sell. Only to realize that within the next six months we are likely to see prices rise again. But hey, maybe he bought it for $2 an ounce 15 years ago and is making huge gains.

That’s 1 kg or 2.2 lbs of coins but I am not going to buy that much because I don’t want to put all my wealth into just one asset. There are other assets and instruments in existence as well.

Related Post

Virtualized Windows GPU passthrough performance

548 frames per second in this particular GPU test.

I got two 1440p monitors for the host Linux machine and then third 1080p monitor for the Windows machine. Then I use Seamless Synergy to share clipboard and mouse/keyboard with that virtualized Windows. So when I move the mouse outside the farthest host monitor, it moves to the Windows monitor over the network.

Pretty neat setup for work. Quite an overkill but because I can I will. Still need to install more SSD to get mirrored pool for redundancy.

The old Radeon 8570 is more than enough for Office and every other application. Can’t quite run 4K video but 1440p plays almost perfectly. On Youtube I see small number of frames being dropped but nothing major.

Caches are causing some funny business:

But it takes about 10 seconds to warm-boot the Windows 10. Meaning that once I have booted it once so that all the necessary data is cached on the Linux host’s memory.

Related Post

Disk failure

Did not take long for one of the cheap 300GB disks to fail.

  pool: pool2
 state: DEGRADED
status: One or more devices is currently being resilvered.  The pool will
        continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
  scan: resilver in progress since Sun Dec 25 13:17:30 2016
    183G scanned out of 612G at 18.3M/s, 6h39m to go
    45.9G resilvered, 29.98% done

        NAME                                             STATE     READ WRITE CKSUM
        pool2                                            DEGRADED     0     0     0
          raidz1-0                                       DEGRADED     0     0     0
            replacing-0                                  OFFLINE      0     0     0
              luks-80ec16d0-3d1f-4ed5-a4f0-0ee564547280  OFFLINE     52   191     0
              luks-fd4c98cc-d1be-4812-ae33-3650fdc984c0  ONLINE       0     0     0  (resilvering)
            luks-e7868ceb-fe11-4e40-a5e3-b6e56129b380    ONLINE       0     0     0
            luks-31f81d2e-36eb-4213-b101-e17d13907df5    ONLINE       0     0     0
            luks-c265d6da-5c40-435b-9f95-c2512c5a8bc5    ONLINE       0     0     0

errors: No known data errors

Had to take the 300GB from another pool until I get a replacement.

zpool offline pool2 luks-80ec16d0-3d1f-4ed5-a4f0-0ee564547280
zpool replace -o ashift=9 pool2 luks-80ec16d0-3d1f-4ed5-a4f0-0ee564547280 luks-fd4c98cc-d1be-4812-ae33-3650fdc984c0

I think I’ll wait for the missing hardware for the additional 8 bays and then migrate from 4x300GB RAIDZ1 to something else. Maybe 6-disk RAIDZ2 with online spare. Better to use those disk that I have. Or maybe get a bunch of small SSD disks from work.

Detecting counterfeit coins

I slowly began to realize that buying the cheapest silver (right on the spot price) could have some problems such as higher likelyhood of buying counterfeit.

So I emailed the local coin dealer and asked them what methods do they use to make sure they are selling the real deal; obviously they haven’t answered yet; it is Christmas time, so I ended youtubing (yes, that is a thing) about the topic and found out that strong magnets can be used to detect counterfeits.

But the videos also showed that for example copper has almost exactly the same magnetic behavior as silver, so simple neodymium test alone could not be used, but weight and volume should also be checked.

But that is a lot of work and the coun dealer obviously wouldn’t be doing that or they would go out of business very fast.

So what else could they be doing?

I believe they use electronic devices to reference the coin’s magnetic properties against mint reference data. Because if you build a device that can accurately measure magnetism and how a coin changes the magnetism within the test apparatus, then it becomes impossible to replicate that magnetic field with anything but the real metal.

I probably won’t be building one but instead leave it to dealer if they give me this sort of answer. They obviously must do something along these lines. Acid or other chemical tests aren’t possible on valuable coins and there aren’t many other feasible methods to accurately distinquish fake coins from real coins.



Related Post

Canadian Maple Leaf, American Eagle or Vienna Philharmonic?

So they are all the same, so which one do you pick? Prices are practically all the same, plus or minus a one USD.

But there’s the catch. American Eagle is 1 USD, legal tender. Vienna Philharmonic is 1,50 €, legal tender. Canadian Maple Leaf is C$5, also legal tender.

So if the prices are the same, plus minus 1 USD, why would you pick USD when the Canadian Maple Leaf is worth $3.50 US. You would if you believe it has better recognition, or if you believe it is of so much better quality, or if you believe the Canadian Dollar is in worse shape than the debt-burdened US dollar.

I don’t know a lot of Canadian Dollar, but I know enough about US dollar and the Euro to know that somethign worth 5 could be more valuable than either of the debt currencies. Of course the Vienna Philharmonic would be legal tender where I live, but if I ever need to buy anything with silver and use it for its non-metal value, then something has gone so horribly wrong that the face value does not matter. Although it would be nice to experiment buying goods with silver, because it is legal tendern, and at least the government probably is required to take it, but honestly I don’t want to risk loosing that much value for a can of coke.