New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Sorry we do not have public listing, as we mostly only look for local people.
You can contact via sales with a open format application tho.
That must've been a really really long time ago, close to 12 years. I believe HTTPS was enabled a few months or weeks after we opened the doors and first customers were setup.
Self signed cert on the other hand has been the norm most of the time, since there was no automated way to get bulk licenses and cost was initially prohibitive a decade ago.
These days if there is self signed cert it's when once again certbot has broken by what-ever 3rd party reason. EFF intentionally breaks it every now and then for old users.
@PulsedMedia thank you for your time explaining to the peanut gallery. I despise the crypto scam but i can certainly understand that if you are building a server for high I/O throughput combining a very bursty high priority load with a low priority background load just makes sense for optimal server use. I also hate capitalism but i get that servers gotta pay the bills, and that its gonna be a more usable server with a background load than 2ce as many burst loads competing.
If people dont wanna benefit from the crypto scam even indirectly as a moral principal i guess i get it, but either way the scam is gonna be running. It isnt gonna change til we regulate, you harming yourself wont change it. And this being the far less energy intensive version, who knows, may even do a tiny good moving people off proof of work systems.
Just restocked with a few big servers
Next servers for this lineup might take until end of january
I have the 450GB plan.
I'm thinking about canceling the HostHatch 250GB KVM, and move the backups here.
They are not movies or ISO's, but family photos (unencrypted) and tax returns (encrypted).
@PulsedMedia any chance the year 1TB flash deal can get sonarr as well?
sonarr is on all plans, just config + launch from shell.
It's trivial to encrypt everything btw, just pipe data through gpg
for example:
That would tar whatever data you want, gpg can do also the compressions and finally passed through ssh to remote server file. or local just remove ssh user@server
unencryption is just gpg decompress and pipe to tar. If you got big data set you can even use pigz -9 to compress with multiple cores and maximum compression level -9
"Family photos" is the new Linux ISO'S. But I give @Daniel15 first credit.
Encrypt his photos? Have you seen his posts? He'd be happy to have some random hacker download and view all those pictures.
am I right, that for your current level of inter-user isolation (or rather lack of it), if users were able to spot Chia harvester process, then they will be able to sneak this above command as well and find that secret "user@server" (the heart of world wide pushup CDN infrastructure) involved in processing yoursunny's precious "family photos"?
EDIT: or maybe that command is meant to be run on client side, so then no issue at all
LOL. I did actually mean it legitimately.
@PulsedMedia i just claimed one of the seedboxes, thank you for what you do!
Thank You for choosing Pulsed Media
Getting low on stock, so get yours now if you are looking to get one.
New nodes i don't think we can get online before mid January.
We just racked a ~rack full of 4x3.5" nodes. Drives, CPUs, RAM, network stuff etc. are already in stock tho. QA, 2 more racks of servers to rack, holidays etc. so might even take until around february new stock starts to rolling in significant numbers.
Dragon-R is sold out, but new mega server is already on order, and another one of those is planned to be ordered by february. First one with direct connection 36x 3.5", and second one with SAS 12G Expander and SAS drives, should be even faster. Both will be 32core EPyC as is normal for the Dragon-R series
These are north of 20 000€ each if maxed out config, but to drive even further performance current plan is to go with 8-12TB drives instead of 18TB. Not your typical seedbox server then, and it's always fun to see GB/s rates read+write simultaenous for 100% random I/O on this level of storage, just insane perf these megaservers put through 
M10G has barely any stock either.
All of the other offers are practically sold out.
1 or 2 server installments might appear here and there until we get the new servers running. Considering also going for atypical config for many of them, and leave the second CPU in. Many of the servers are 2x12Core Opterons. Would make for wicked compute servers @ budget so we might just offer them as dedis.
Ok..
What configuration are you using with these? Hardware RAID? Linux md raid? bcache? Chunk size for the raid set?
Curious minds would love to know
These will mostly be RAID0 V10G servers. typically something along the lines of:
mdadm --create /dev/md1 -l0 -n4 --chunk=2048 /dev/sd[abcd]4There is currently reaaaally bad congestion on Twelve99/Telia network: https://pulsedmedia.com/clients/announcements.php?id=537
Hopefully resolved by tomorrow.
Thx a lot for already implementing qBittorrent
I just gave it a try on a V1k L instance and it runs pretty well (with only a few torrents so far). Although in direct comparision with rtorrent i've seen some >150% cpu usage when reaching max. download speed download (1gbps) from some peers.
Speaking of recent implementation is there a special reason it's still on v4.1.5?
Seems this is from Dec. 24th 2018 which is even older than last 0.9.8 rtorrent rls from 2019 😇
Needless to say qBT got many improvements over the last years but personally to fully replace rtorrent for me I would love at least v4.3.0 release for:
FEATURE: Add RSS functionality in Web UI (Sepro)There are also alternative WebUIs worth checking, e.g. https://flood.js.org/ (as it's a single super sleek UI supporting rTorrent, qBt, Transmission and Deluge experimental as backend)
Got it up for testing in a few minutes but don't wanna run anything "unofficially" on the server of course.
Cheers!
Good thing seedbox servers generally have their CPUs almost fully idle then, ain't it?
We currently use the repo version of qbittorrent. No sense for us to go through the trouble of maintaining from source until we see it gets enough usage etc. and a solid reason for that.
Flood has been requested by many indeed.
You can run it "unofficially" no problem, we cannot have time to implement everything under the sun, so there is no "unofficial" usage. Just be mindful of other users, that's all
Mine-s working great. Thanks @PulsedMedia
@PulsedMedia any ETA for the official qBittorrent rollout on the M1000 A2 series?
It would be godsend!
I always hated that rutorrent starts to throw timeouts all the time when there are ~2800+ torrents.
Nice, thank you kindly for that relaxed point of view
Regading the version, oldstable Debian Repo is really pretty conservative 🙈
There is also a Git Repo constantly publishing static linked binaries of latest Qbt, running great out of the box 👌
https://github.com/userdocs/qbittorrent-nox-static
Implementation would be as simple as:
wget -qO ~/bin/qbittorrent-nox https://github.com/userdocs/qbittorrent-nox-static/releases/latest/download/x86_64-qbittorrent-noxI see 6 cores and 48GB RAM.
Since it's 24 users per server, everyone gets 25% core and 2GB RAM, right?
However, I can't figure out how to set CPU and RAM limits.
Last time I tried the pre-installed ffmpeg, it goes up to 200% CPU, so I quickly stopped it to not ruffle any feathers.
The two methods I know are Docker and systemd, but neither is possible without root.
As for running programs unofficially, one issue is port number allocation.
I notice that uTorrent and lighttpd are each assigned a port number whose last two digits are same as account UID.
Does it mean I can use any port number that end with account UID?
Is there any other guidelines I should know about?
Then, I see I could insert custom lighttpd config, but how to make lighttpd reload the config?
If lighttpd config change could take effect, there would be no port number problem because I can then have the app listen on Unix socket and use lighttpd as reverse proxy.
It is already rolling out on all. If not on your server open a ticket requesting your server to be updated.
rutorrent works with 14k+ torrents on our setup, if not then that server is busy and that comes with the territory of entry level shared.
If that's the same userdocs who was mod on reddit, like i think he is, we will not touch anything he does or has done. He will cause a shit storm just to spite us. You would not believe the shit he tried to pull of in his moderator power trip.
Not a VPS, so you can burst use all those 6cores.
ffmpeg if i recall right has a flag for how many cores to use, and just use nice 19 to start it, then it does not matter if you even try to use all cores as only idle cycles would be used.
port numbers ending on UID is purely coincidental.
restart lighttpd: Kill the process, and automation will start it again within 2 minutes -- or you can start it yourself
I can never imagine how someone could mentally keep track of 14K torrents.
I have 10 movies and I'm already confused on which movies I haven't watched.
I know I can delete the movie right after I've watched it, but I read on a magazine that I'm supposed to keep the files for a few weeks before deleting.
I see, cron script or the likes.
I didn't find systemd service, so that I was too afraid to kill lighttpd because I worried it wouldn't come up again until who knows how long.
ffmpeg would become nice 19 after a minute.
Automation strikes again.
You can pull up random tweets from last decade but not recall 10 movies you've watched in last months? Are you sure you're downloading movies worth watching in the first place? Remembering if you saw a shitty movie likely is harder to remember.
Remembering the last 14k movies I may have watched is tricky, not the last 10 in a list.
If we didn't have a watchdog we'd be swamped with tickets X)
It's actually kinda ridiculous how many watchdogs we need to have, and how flaky most software actually is.
Automation strikes again.
Yup, i think i forgot we might have automation for that too.... oh well, so much stuff happens in more than a decade cannot recall every single detail
@PulsedMedia My service is down. 502 Bad Gateway on website, and Torrent trackers are showing no activity.