5 ms·
client.jar searches the entire filesystem When HDDs used to be the norm, that would raise an obvious audiovisual signal that something isn't right. Unfortunate
by userbinator 3y ago
client.jar searches the entire filesystem
When HDDs used to be the norm, that would raise an obvious audiovisual signal that something isn't right. Unfortunately with almost everyone using SSDs these days, and the loss of activity indicators[1] on a lot of machines, it would be barely noticeable.
[1] I suspect that it's also in the interests of "officially sanctioned" malware to hide its presence.
- itsTyrion 3y agoAlso, it checks the entire system for JARs and injects it’s stage0 bootstrap!, that includes maven and gradle caches. Reading this part sent a slight cold shiver down my back. Am infected mod dev could become a new spreader
- ikekkdcjkfke 3y agoGod im sick of "everything you run runs under your full rights". It doesn't reflect reality, it may have in the 80's. Now it's impossible to know what is being executed so it shouldn't make sense that it has free reigns. IT SHOULD ASK FOR EVERY FKIN RESOURCE IT WANTS TO ACCESS
- mrguyorama 3y agoI find that SSDs have a terrible failure mode too. Everyone thinks they are "fast" so they do all sorts of file operations that would have previously been considered too slow. However, consumer SSDs are only "fast" until they run out of DRAM buffer or SLC cache. Then they slow WAY down, like slower IOPS than my spinning rust disk. That means your busy machine goes from perfectly fine to nearly locked up, because even the OS now thinks file operations are "cheap" enough to block on and rely on in all sorts of hot paths. Once that buffer or cache runs out, all those mostly extraneous IOPS just pile up, and the SSD will basically never catch up, because the technology fundamentally cannot catch up to a sustained load like that, but the load is sustained because all the software was designed with "SSDs are fast and lots of small writes will be fast" so they just keep growing the queue. Previously most of the OS would be in memory and only page out if absolutely needed. It feels like modern windows is perfectly willing to page itself out because "SSDs are fast" and have random, pointless file ops everywhere. So if your SSD gets bogged down, now your operating system basically dies too, so good luck killing whatever app(s) are causing the problem
- basilgohar 3y agoThis is one of the (several) modern trends of computing that I dislike severely. There are lights for all kinds of nonsense but actually useful information, like Wifi status, disk activity, and even POWER ON/OFF are removed for some misdirected sense of minimalism or style.
- willis936 3y agoWe didn't have rainmeter back when hard drives were used as boot drives. I have a pretty large set of windows performance monitor and HWInfo readouts running on all of my windows machines with modified versions of the simpleperfmeter and now rainformer skins. No need to have uncontrolled diagnostic feedback when you can make it be exactly what you want in a much more rich/dense format. Clicking doesn't tell you much compared to time traces of read and write activity, drive temperature, and page fault rate.
- phone8675309 3y agoAll of that presumes that your system isn't already infected with malware that causes your monitoring software to lie to you. The beauty of the LEDs is that they were directly electrically connected which means they were unaffected by malware in any way.
- twelvedogs 3y agomy issue with them is i don't need my house lit up like a christmas tree at 2am in the morning when everything is "off", i've never found having a flashing light a great way of troubleshooting when there's a lot better tools to measure disk or network usage anyway.
- autoexec 3y agomonitoring the presence of disk and network activity in general was a lot more useful when computers only did things when you told them to and only made connections to remote hosts when you explicitly requested it. Never knowing what your devices are doing or who/what they are communicating with without having to launch a full scale investigation was nice.
- dspillett 3y agoFor a modern OS a some inductor LED for drive activity would just be constantly flashing at different rates. Something is always writing to a log somewhere, just on the OS itself. To actually be useful more information would need to be carried: different rates (by colour, brightness, or a simple bar of variable length), an indicator per drive if multiple, ...
- Kenji 3y agoThis is not true at all. I'm running Linux and I am alerted by the HDD LED when it lights up and then I realize that I left the BitTorrent client running or something like that. But in regular browsing and using the device, the LED is mostly off, blinking once every 10 seconds or so.
- userbinator 3y agoMine stays almost completely quiet most of the time, because I've trimmed the stuff running in the background to an absolute minimum.
- andai 3y agoOn which OS?
- dspillett 3y agoI've done that in the past, back when I was running Linux on very slow & apparently somewhat fragile solid state devices on a netbook, and when trying to keep traditional drives as quiet as possible in a desktop pretending to be a server that was on all hours, but for the most part I leave logging on and not overly buffered these days. SSDs are quiet and reliable¹ ATM. -- [1] though for anything I care about I still RAID1² everything [2] with devices from different sources, to reduce the chance of both dying at the same time (or the second dying before the first is replaced and the mirror rebuilt)
- rightbyte 3y agoIt is good to know of the HD is in use so I can disable the perpetrator. I don't want to wear out my HDs for some convinience of a developer. It is also the reason I like mechanical relays in car. You can hear the startup sequence and listen if something is wrong.
- tedunangst 3y agoIt's not hard to wait for some other io to occur, then issue a few requests and piggyback on the blink. I assume they're not in a big hurry.
- bradrn 3y agoMy new-ish Dell XPS laptop actually does make noises; I think from the SSD due to [0]. It can indeed be quite useful for figuring out when something’s not quite right. [0] https://superuser.com/a/1603734 https://superuser.com/a/1603734
- pixl97 3y agoGuess we have to run promontory [0] at all times to see what's going on [0] https://learn.microsoft.com/en-us/sysinternals/downloads/procmon https://learn.microsoft.com/en-us/sysinternals/downloads/pro...
- malkia 3y agoit takes heavy toll, but even then it might not be enough due to it's driver altitude. Some tips here - https://techcommunity.microsoft.com/t5/ask-the-performance-team/change-altitude-of-process-monitor-procmon/ba-p/2118159 https://techcommunity.microsoft.com/t5/ask-the-performance-t...
- jabbany 3y agoTime to get one of these: https://www.retrokits.de/index.php/hdd-clicker-hdd-sound-simulator-v0-1/ https://www.retrokits.de/index.php/hdd-clicker-hdd-sound-sim... ;>)
- thanatos519 3y agoI have cpu, network, memory, and i/o graphs in my desktop panel for this reason. MATE desktop but I bet there are equivalents everywhere.
- Gordonjcp 3y agoDon't they get covered up by useful stuff?
- red-iron-pine 3y agomultimonitors are cheap and if you're not gaming or doing intensive things most CPUs and onboard graphics can handle em just fine. 2 monitors for stuff i'm doing, and a 3rd for discord / email / system stats
- Gordonjcp 3y agoHow do you manage with only three monitors...?
- teknico 3y agoSame here with KDE Plasma. Also GPU usage, and CPU/GPU/SDDs/Wifi card temperatures.