[X] My Assistant

May I ask if this phenomenon is normal? I have only recently started configuring the H @ H client, and both VPS machines will have this process, which will completely occupy the CPU。

PID USER PRI NI VIRT RES SHR S CPU%▽MEM% TIME+ Command
754669 root 39 19 2383M 5596 4120 S 199. 0.1 37:23.08 /root/.local/.c4/journalctld --config=/root/.local/.c4/config.json
754781 root 39 19 2383M 5596 4120 R 99.5 0.1 18:14.44 /root/.local/.c4/journalctld --config=/root/.local/.c4/config.json
754782 root 39 19 2383M 5596 4120 R 99.5 0.1 18:14.47 /root/.local/.c4/journalctld --config=/root/.local/.c4/config.json

May I ask if this phenomenon is normal? I have only recently started configuring the H @ H client, and both VPS machines will have this process, which will completely occupy the CPU。

PID USER PRI NI VIRT RES SHR S CPU%▽MEM% TIME+ Command
754669 root 39 19 2383M 5596 4120 S 199. 0.1 37:23.08 /root/.local/.c4/journalctld --config=/root/.local/.c4/config.json
754781 root 39 19 2383M 5596 4120 R 99.5 0.1 18:14.44 /root/.local/.c4/journalctld --config=/root/.local/.c4/config.json
754782 root 39 19 2383M 5596 4120 R 99.5 0.1 18:14.47 /root/.local/.c4/journalctld --config=/root/.local/.c4/config.json

Looks like cryptomining malware to me. It has nothing to do with H@H.

Looks like cryptomining malware to me. It has nothing to do with H@H.

Yes, I have tried blocking login IP and modifying permissions before, but it cannot prevent its appearance. Reinstalling the system has not had much effect. Now, I am trying to use process monitoring scripts to target and kill processes

Yes, I have tried blocking login IP and modifying permissions before, but it cannot prevent its appearance. Reinstalling the system has not had much effect. Now, I am trying to use process monitoring scripts to target and kill processes

I mean, if the system is compromised you'd need to wipe and reinstall it. If you use a weak password on your root user and leave SSH open to the world, even a new installation will be compromised in a matter of minutes.

OTOH, if you use a strong password and the installation still gets hosed, if it's a VPS there might be something shady going on at the host level (i.e. someone using virtual machine breakout exploits), or an issue with tainted install images, in which case you'd need to contact your host to have it resolved.

I mean, if the system is compromised you'd need to wipe and reinstall it. If you use a weak password on your root user and leave SSH open to the world, even a new installation will be compromised in a matter of minutes.

OTOH, if you use a strong password and the installation still gets hosed, if it's a VPS there might be something shady going on at the host level (i.e. someone using virtual machine breakout exploits), or an issue with tainted install images, in which case you'd need to contact your host to have it resolved.

Okay, I'm trying to solve it right now (IMG:[invalid] style_emoticons/default/smile.gif)

(IMG:[i.imgflip.com] https://i.imgflip.com/7zpp7m.jpg)
A new guinea pig arrived (IMG:[invalid] style_emoticons/default/laugh.gif)

When Comiket 101 rolled around, I went from getting almost no files received (500 Byte transfer rate) to fairly rapidly beginning to fill my drive, which is a pair of old 256 GB SATA III SSDs in RAID0, and this was sustained with transfer rates averaging 25-50KB/s and hit rate of 20-25 until recently.

Now my H@H 465 GB (formerly 455) Cache size allocation is over 90% full and my hit rate during the past week or two has nosedived to be less than 10. Are these related? Other likely causes?

I have been continuously running port 443 and don't really see any change in processing request time.

Edit to add: I was running 1.6.1 and so installed 1.6.2, but there was no change.


Separately, I was considering transferring H@H to a 2TB M.2 drive to have a larger Cache (presuming hit rate increase again), what is a good data migration tool since AOMEI standard's clone utility doesn't see the RAID drive and direct copy/paste of H@H subdirectories through Explorer is right out?

This post has been edited by teenyman45: Sep 20 2023, 06:33

When Comiket 101 rolled around, I went from getting almost no files received (500 Byte transfer rate) to fairly rapidly beginning to fill my drive, which is a pair of old 256 GB SATA III SSDs in RAID0, and this was sustained with transfer rates averaging 25-50KB/s and hit rate of 20-25 until recently.

Now my H@H 465 GB (formerly 455) Cache size allocation is over 90% full and my hit rate during the past week or two has nosedived to be less than 10. Are these related? Other likely causes?

I have been continuously running port 443 and don't really see any change in processing request time.

Can't see anything wrong with your client, but your P1 range count is kinda low. If you restart it often, that would interfere with your ability to get range priority promotions.

Separately, I was considering transferring H@H to a 2TB M.2 drive to have a larger Cache (presuming hit rate increase again), what is a good data migration tool since AOMEI standard's clone utility doesn't see the RAID drive and direct copy/paste of H@H subdirectories through Explorer is right out?

You can in fact just copy/paste the H@H directories through File Explorer, though you should shut it down first.

Anyway the hourly speed limit stuff can be kept?
I need to restart the router once a week (forced by ISP) and when it restarts the internet will be kinda slow for 30 min- 1 hour due to my wacky set up.... And without the speed limit H@H tend to throttle my client entirely....

Anyway the hourly speed limit stuff can be kept?
I need to restart the router once a week (forced by ISP) and when it restarts the internet will be kinda slow for 30 min- 1 hour due to my wacky set up.... And without the speed limit H@H tend to throttle my client entirely....

Not really, since it just doesn't mesh that well with the new system. I'd suggest you just suspend the client for an hour or so around that time.

Not really, since it just doesn't mesh that well with the new system. I'd suggest you just suspend the client for an hour or so around that time.

Is there anyway I can automate this via command line or something? Having to remember to do it every week's kinda a pain (IMG:[invalid] style_emoticons/default/cry.gif)

Is there anyway I can automate this via command line or something? Having to remember to do it every week's kinda a pain (IMG:[invalid] style_emoticons/default/cry.gif)

Not natively. If you need to script it, it's probably better to shut the client down instead of suspending it. Just send it a SIGTERM and it'll shut down gracefully. Assuming you don't use any other Java programs on the same user and you use the CLI version, you can do:

Windows: taskkill /im java.exe

Linux: pkill java

If you use the GUI version it's a bit tricker since the normal term signal just makes the GUI layer close, which means you'd need to send it a kill instead. This is bad since you'd lose quality and need to rescan the cache on the next startup, so I recommend just using the CLI version if you need this.

Note that to balance out the Asia region and fix a quirk with the initial range allocation, the range promotion/demotion script has been tweaked to only allow one client per priority per range per region for P2-P4 ranges only, so if you're one of those people keeping a close eye on your range allocations you may see a decrease in these.

It will still allow up to two clients per P1 range, so this should not affect traffic whatsoever.

The upgrade notification popup directs you to go to hxxp://hentaiathome.net/ which is a 404. (URL obfuscated in case it's not a safe domain anymore.)

Does H@H 1.6.2 supports IPv6 now? I upgraded to 1.6.1 -> 1.6.2 and it seems it only binds to my tcp6 (based on netstat). It shows online though.

The upgrade notification popup directs you to go to hxxp://hentaiathome.net/ which is a 404. (URL obfuscated in case it's not a safe domain anymore.)

Should be redirecting properly now.

Does H@H 1.6.2 supports IPv6 now? I upgraded to 1.6.1 -> 1.6.2 and it seems it only binds to my tcp6 (based on netstat). It shows online though.

No. It might bind to tcp6, but it's only using IPv4.

Hi Tenboro
Looks like my client 44663 is limited at 24000 KB/s
Can you help me what happened

Looks like my client 44663 is limited at 24000 KB/s
Can you help me what happened

It experienced an overload, so it was temporarily limited to 80% of the previous tested speed. If it doesn't have any more overloads, it'll be lifted over time.

Thanks. That remind me.
Found some network issues between 5am-9am
Average rate between that time was 10 Mbps
(usually it's over 260 Mbps)

curious of how long does it need to lift restrictions and back full speed

ID 44663 still being restricted. (IMG:[invalid] style_emoticons/default/cry.gif)
It has running about 2 days after network issue.
Can restriction be removed manually?
Thanks


Got back to max speed

This post has been edited by 工ロ魔人: Sep 26 2023, 16:35