I thought you installed Docker to a HDD on the USB, so never mind, I understand now.
examples on scripting that live outside the eco system?
Not sure what eco system you mean, but there's that tutorial link.
I thought you installed Docker to a HDD on the USB, so never mind, I understand now.
examples on scripting that live outside the eco system?
Not sure what eco system you mean, but there's that tutorial link.
Try umount -f -l. You're probably running a shaky USB HDD dock, because all of them are.
It's surprising that you know what to do but haven't already put this is a script, although priority 1 should be getting the Docker stuff off that USB dock.
https://linuxconfig.org/bash-scripting-tutorial-for-beginners
That's EXTREMELY intrusive, are you sure you weren't phished?
But does this mean that SATA...
Does it matter?
GUESSING: It looks like one of the SATA devices of the JMB585 is wired up to the Intel SATA interface for expansion. Your Intel "SATA 2" device seems to actually be the first device on the JMB585. I'm not sure how SATA expansion works, but if it's like SAS, then it increases a 16bit number for addressing. Or, It could also mean that it has to squeeze all 5 devices through that single wired up device. You'll have to RTFM this if you really care for an answer.
PCIe is full duplex so I wouldn't worry about speeds with non-SAS 3.5" spinning rust (eg. Seagate Iron Wolf Pro). Also, the exact physical device correlated to which device /devices/...sdX may jump around based on probing, repeatable assignment is not guaranteed (ie. .../sdc may later be .../sdf).
I've been running Gitea for a few years but the amount of times I've used the GUI is so ridiculously low that I haven't installed it on the last 2 VPS's and I don't miss it a bit. Granted it's a personal Git that I cleaned recently down to about about 4GB in total, but that also includes a bunch of CAD binaries/archives.
FWIW, Gitea seemingly used less RAM than Gitlab.
Who owns $GITLAB_HOME?
Post the error log.
At least this i3 board has a open PCIe x1 header, the Asus board requires adapters for > x1.
Usenet has 50 episodes available in nf webrip 2160p, ~107GB
"internal" in this sense means internal to the package, which in this sense is true.
To "cleaner.py", dead for 1 second is the same as dead for 1 year.
Christmas trolls
This sends *.torrent files and magnet urls to whatever $ip:port qbittorent is on. It did work, but I never really used it but maybe someone can use it as an example. I'm posting because I forgot about it until another thread topic.
#!/bin/bash
function magnet_forward() (
# used with qbittorrent's webserver to add magnet links or files to the queue
# requires Bash and cURL
# the user and pass of the web login
username="admin"
password="adminadmin"
# the ip and port that qbittorrent is listening on
ip_and_port=127.0.0.1:4600;
# enforce ".torrent" to be the last 8 characters of a filename
# 0 = no, 1 = yes
enforce_name=1;
# move torrent files to this location (only torrent files, not URLs)
# - does not overwite, uses: mv -n "$source_file" "$torrents_dir"
# - must be an existing directory
# torrents_dir="/home/$USER/torrents";
if [ "$1" == "no_enforce" ]; then
enforce_name=0;
shift;
fi
if [ ! "$#" -gt "0" ]; then
echo "ERROR: requires at least 1 argument, received $#"
return 1;
fi
# NOTE: you only need to get the "SID" ("Session ID") once each time the
# script runs, not for every command.
# get the "SID" string
IFS='=;' read -ra sid <<< $(curl -v -d "username=$username&password=$password" -X POST $ip_and_port/api/v2/auth/login 2>&1 | grep "SID");
if [ -z "$sid" ] || [ -z "${sid[1]}" ]; then
echo "ERROR: could not parse SID";
return 1;
fi
sid="${sid[1]}";
if [ ! -z "$(echo $sid | grep '\s\|;\|,')" ]; then
echo "ERROR: cookie value contains spaces, a semi-colon or a comma. SID="$(echo $sid | grep '\s\|;\|,')
return 1;
fi
files=();
links=();
for i in "$@"; do
if [ ! -z "$i" ]; then
if [ ! -f "$i" ]; then
# if it's not a file then assume it's a URL.
links+=(-F "urls=$i");
else
if [ "$enforce_name" ]; then
tmp="${i,,}"
if [ "${tmp: -8}" != ".torrent" ]; then
continue;
fi
fi
links+=(-F "torrents=@$i");
# for now, skip the directory check for $torrents_dir
if [ ! -z "$torrents_dir" ]; then
files+=("$i");
fi
fi
fi
done
if [ "${#links[@]}" -gt "0" ]; then
curl -v -H "Cookie: SID=$sid" -H 'User-Agent: Fiddler' "${links[@]}" $ip_and_port/api/v2/torrents/add;
# need error handling, currently assuming success by always returning 0
if [ ! -z "$torrents_dir" ] && [ -d "$torrents_dir" ]; then
for i in "${files[@]}"; do
mv -n "$i" "$torrents_dir"
done
fi
return 0;
fi
return 1;
)
(return 0 2>/dev/null) || { magnet_forward "$@"; return $?; };
# below is a better parser to get the SID, but requires a grep with -P support.
# sid=$(curl -v -d "username=${username}&password=${password}" -X POST $ip_and_port/api/v2/auth/login 2>&1 | grep -o -P '(?<=SID=)[^;]*');
Alles anzeigen
You can pause any modern torrent client via it's CLI or using curl/wget/etc.
With qbittorent it would be something like this, although I don't know the pause syntax (RTFM'n it)
sid=$(curl -v -d "username=${username}&password=${password}" -X POST $ip_and_port/api/v2/auth/login 2>&1 | grep -o -P '(?<=SID=)[^;]*');
curl -v -H "Cookie: SID=$sid" -H 'User-Agent: whatever' "${links[@]}" $ip_and_port/api/v2/torrents/add;
You'll replace this: ${links[@]}" $ip_and_port/api/v2/torrents/add;
... with whatever the pause/resume syntax (assuming you're using http to communicate).
You've seemed to completely ignore my suggestion about scrub cancel/resume, so I assume you didn't try :-/
boot either one via BIOS selection at boot. But it didn't work
Manually, by human choice? Pretend it can be made to work, what's the point?
At the "Enterprise" level, which is what a lot of these tools are designed for, all data and people are replaceable, remember that corporate view.
Is your data irreplaceable? Every abstraction avoided helps resist munging so, if it's irreplaceable data, you probably don't want containers or VMs or even to have the drive powered on!
No nothing like that, never mind me. Just roll with it.
The X550AT2 can be had in PCIe 3.0 x4, but sadly I think the x4 is only RJ-45. Still, sometimes it's desirable to have full speed at x4 instead of x8 (ie. break out a NVMe)
Ah, I see you chose 1 port at x4. FWIW, everything RJ-45 based is expensive in 10gbe+, you might want to look at SFP.
Based on your 1st post
mobo: https://www.ebay.com/itm/265169958785
RAM: https://www.ebay.com/itm/256281740887
Mobo + CPU + RAM before tax in USD: 300.91
I had that board and the ground for the RJ-45 10gbe port was seemingly picky, I switched to a S/FTP cable and all was well. 6 SATA ports are available. The PCIe 3.0 x16 bifurcates to x8x8.
IMO older used Intel Xeon D or Intel Atom boards are your best for dollars-to-features which will support bog standard ECC RDIMM (not the lesser LRDIMM or unbuffered). Anything from D-15xx or Atom C3xxx will do.