memory hardware – A doubt on best media for a time capsule

So, I have some data to a time capsule project of mine. But the thing is, I need to know the best media on long-time scale to use as storage unity. The time range is 120 years. Now, I’ve been thinking about two medias: the average CD-ROM/DVD-ROM and a outdated Vinyl/LP. On the one hand the CD/DVD is a good one, due to its dimensions and information space, but the question remains about the difficulties to read a CD/DVD in 2141. A vinyl LP could be more simple.

So my question is: what is the best media for a time capsule?

c++ – How does boost sort handle additional memory allocation

According to the documentation of boost::sort (https://www.boost.org/doc/libs/1_75_0/libs/sort/doc/html/index.html) all algorithms uses “additional memory“.

I couldn’t find any information on how this memory is allocated in the documentation.

Does anyone know if it is heap-allocated during each sort operation, is it stack-space, or does boost::sort utilize some static internal memory buffer (like std::stable_sort)?

If heap-allocated, is it possible to pass a pre-allocated buffer to the sort operations in order to gain control over memory allocations?

Windows 10 v1909 Memory Leak

Windows 10 Pro v1909 Intel i5-8250U 8GB RAM

I updated to v1909 from v1809 back in December, and since then, I’ve noticed that my Memory used at idle and my paged/non-paged pool is significantly more than it used to be, particularly after my computer has been running for awhile. This problem was present in a somewhat more mild form in 1809, but not at all in 1803.

Prior to the update, my system would take about 21-30 days to grow from 100mb paged/400mb nonpaged to ~500/1000 respectively, in about 2 weeks or less (most recently in just over 9 days), assuming I just do basic web browsing; some YouTube, some Twitch at most. If I do things like run my antivirus, or a decent bit of troubleshooting (eg lots of opening new windows), it seems to grow faster. I don’t recall what my paged/non-paged used to be, but I keep seeing that ~300mb is supposed to be normal for non-paged, and mine is usually about 600-750mb, but can sit at 1GB after only 14-16 days uptime. Neither seem to be ticking down at all when left alone; if anything, they seem to very slowly be ticking up, indefinitely.

I’ve been trying to figure out the source of the change/problem, but I’m not exactly computer savvy; I just have Google and some spare time.

I tried using PoolMon, and from what I could make sense of, the largest non-paged thing listed was IAStor, which I do not know where to update the driver for (if it has one). Also, the Empty Standby List thing in Rammap didn’t work, either. I have also tried restarting Windows Explorer after seeing that it had a lot of paged/non-paged memory usage, but it did nothing.

I’ve also tried to see if it was something to do with my Network Data Usage Monitor (I keep seeing it referred to as NDU), as I’ve seen that be listed as a culprit (though usually with Killer as the device, whereas I have Intel as best I can see) particularly since the “Network” box has a habit of disappearing from my Task Manager’s Performance tab (it also doesn’t show usage in Processes either, even though the status column is there) after my computer has been running for awhile. I ran the diagnostic for it in Windows, and it didn’t see anything. Network still doesn’t show in my Performance tab so I don’t think it did anything, and I don’t know what else to do to pursue this avenue.

Has anyone else had this issue? Is it some sort of driver issue or is it a Windows issue? And more importantly, does anyone have a clue as to how I can resolve it (preferably without having to trapeze through the messy business of altering the registry)?

windows – How can I find the root cause of nonpaged pool memory leak from pool tag PdcA called by umpdc.dll?

I have a nonpaged pool memory leak that grows consistently with about 80 allocations per second, each 200 bytes. So it is about 16kB per second or 58MB / hour, 1.4GB/day.

I have run poolmon.exe and sorted by bytes and can see that pool tag PdcA ha consistent positive change of nonpaged memory about 80,000 bytes per 5 second refresh.

I ran xperf and analyzed the pool.etl file in Windows Performance Analyzer and can see that the stack shows umpdc.dll!PdcpAlpcCallback being responsible for the growth with impacting size matching the growth rate seen in poolmon.exe (see attached screenshot). WPA Screenshot showing 160kB increase in 10 seconds

I also turned on the nonpaged memory column in the task manager details tab and saw about 600 MB of nonpaged memory associated with audiodg.exe (poolmon.exe shows about 6.5GB associated with PdcA). When I killed audiodg.exe from task manager, I saw a few MB of negative change to the PdcA nonpaged pool bytes, and the continuous 80KB/5 second growth stopped and went to 0, but about 6.5GB of memory is still unfreed in PdcA.

I’m sure a restart will free that 6.5GB, and it seems clear that audiodg.exe is the culprit, but I understand that is a standard windows component. The call stack in pool.etl doesn’t show anything specific that is the root cause. And it doesn’t even show audiodg.exe, but I know that to be the cause because killing that process stopped the memory growth immediately. I’m guessing audiodg.exe set up the callback that is being executed (umpdc.dll!PdcpAlpcCallback)? How can one find out what other software (if any) triggered audiodg.exe to set up that callback in the first place? Why doesn’t more information show in the call stack in WPA?

How Can I find out how to stop this from continuing without disabling windows audio support that audiodg.exe provides? I don’t have any audio devices attached other than plug and play hdmi monitor with built in speakers. I am not running any audio specific software other than an occasional web browser with a video that plays audio.

I am running Windows 10 version 20H2 build 19042.804

Thank you for your help.

rhel7 – Apache : socache_shmcb:error AH00820: shared memory segment too small

For the context : Apache HTTPD 2.4.6-90.el7 on RHEL 7.4

  • I had a AH01873: Init: Session Cache is not configured (hint: SSLSessionCache) warning in the logs, so I added to the configuration :
<IfModule socache_shmcb_module>
    SSLSessionCache shmcb:/run/httpd/sslcache(512000)
</IfModule>
  • I checked the socache_shmcb_module module is loaded :
$ httpd -M | grep shmcb
socache_shmcb_module (shared)
  • I checked the syntax was fine :
$httpd -t
Syntax OK
  • then did a graceful restart :
httpd -k graceful
  • afterwards, I saw no more AH01873: Init: Session Cache is not configured (hint: SSLSessionCache) warning in the logs. Hours later, the whole HTTPD server was not functional anymore and the logs reported :
(socache_shmcb:error) AH00820: shared memory segment too small

The Apache documentation itself gives no details about the size of this cache, the value I applied is actually the one shown in examples.

How may I determine the acceptable size of this cache ? Do you have any details regarding this AH00820 error ?

bash – convert-im6.q16: memory allocation failed for converting from gif to bmp

I am trying to convert a file from .gif to .bmp using this command:

convert -coalesce pixels8.gif out.bmp

While doing this i get this error:

convert-im6.q16: memory allocation failed `pixels8.gif' @ error/gif.c/ReadGIFImage/1303.
convert-im6.q16: no images defined `out.bmp' @ error/convert.c/ConvertImageCommand/3258.

I am not sure what the problem is. I have taken this command from https://github.com/glouw/paperview. My system has 8GB ram.

pixels8.gif
Type: ‘image/gif’
Size: 4.6 MB (46,42,849 bytes)
Resolution: 1920×1080

insufficient memory – ninja failed with: exit status 1

I am building an android code on a Linux machine using the following commands but it seems to keep failing for unknown reasons:

source build/envsetup.sh
lunch sdm845_userdebug
./build.sh
16:50:42 ninja failed with: exit status 1

#### failed to build some targets (18:28 (mm:ss)) ####


real    18m28.675s
user    62m57.737s
sys 3m4.666s
============================================
(build.sh): FAILED: make  ENABLE_AB=true BOARD_DYNAMIC_PARTITION_ENABLE=false SHIPPING_API_LEVEL=28

From what I looked up, it could be possibly related to the memory but it doesn’t look like it.

              total        used        free      shared  buff/cache   available
Mem:            15G        1.5G         11G         19M        2.2G         13G
Swap:          975M        855M        120M