![]() |
|
|
#89 |
|
Aug 2002
7×1,237 Posts |
We have set up a dynamic DNS IP address forwarding thingie so others can log into the system (SSH) to test things.
If you are interested in this just let us know via PM.
|
|
|
|
|
|
#90 |
|
Aug 2002
100001110100112 Posts |
Here is a picture of the system all set up.
On the lower left is a small (450VA) UPS and our ancient wireless keyboard/mouse combo. The LCD is also ancient but it works. We were going to do everything headless but we had the LCD so why not use it? We sit very close to this so we can see the temperature readout easily. We ended up using the onboard Wifi for network connectivity. The UPS is severely undersized but it should be enough to get through the small power blips we have out here in the middle of nowhere.
|
|
|
|
|
|
#92 |
|
Aug 2002
865910 Posts |
The antennas are hanging out the back, off of the I/O plate on the motherboard.
We haven't done any special testing other than to make sure the connection is up. The downside to using WiFi is we have to use a non-free package. https://packages.debian.org/buster/firmware-iwlwifi Code:
m@tr:~$ vrms
Non-free packages installed on tr
firmware-iwlwifi Binary firmware for Intel Wireless cards
1 non-free packages, 0.1% of 1506 installed packages.
|
|
|
|
|
|
#93 | |
|
"6800 descendent"
Feb 2005
Colorado
13538 Posts |
Quote:
|
|
|
|
|
|
|
#94 |
|
"Curtis"
Feb 2005
Riverside, CA
2·2,927 Posts |
I'm no security expert, but I've found fail2ban a very useful (if not necessary) program for my internet-facing machines.
An aliquot sequence or two are interesting ways to gain experience with factoring tools. You can automate every step with yafu and aliqueit, or have aliqueit exit every time a composite above some threshold survives ECM so that you can play with parameters & programs more manually. Also, this thread encouraged me to explore upgrade options for my own hardware, where I [re]discovered that X99 boards that support the old 6-core Haswell chips also support Xeons of that era. So, $150 later, my 5820 i7 is now a Xeon 2680v3 12 core@2.5ghz and my most powerful matrix-solving chip. Last fiddled with by VBCurtis on 2020-01-10 at 17:08 |
|
|
|
|
|
#96 |
|
"Mihai Preda"
Apr 2015
22·3·112 Posts |
|
|
|
|
|
|
#97 |
|
"Curtis"
Feb 2005
Riverside, CA
10110110111102 Posts |
Ebay. The seller's name was "pccompd".
Here's the listing (dunno how permanent such a link is): https://www.ebay.com/itm/193283688016 |
|
|
|
|
|
#98 | |
|
Aug 2002
865910 Posts |
Quote:
seq 5 | parallel -j 2 -n 0 nice -19 ecm -inp c1133.txt -timestamp 1e9 | tee -a results.txt This example runs the command five times and it sends the output to the display and appends the output to a file. Parallel schedules all of the jobs, in this case using a maximum of two threads at a time. Note that it is very easy to run out of memory doing this!
|
|
|
|
|
|
|
#99 |
|
Aug 2002
7·1,237 Posts |
Even better with real-time progress information!
seq 5 | parallel -j 2 -n 0 --progress nice -19 ecm -inp c1133.txt -timestamp 1e9 >> results.txt Silly example: Code:
$ seq 1000 | parallel -j 10 -n 0 --progress nice -19 ecm -inp c1133.txt -one -timestamp 1e5 >> results.txt Computers / CPU cores / Max jobs to run 1:local / 24 / 10 Computer:jobs running/jobs completed/%of started jobs/Average seconds to complete local:10/500/100%/0.3s
|
|
|
|