[X] My Assistant

Nice (IMG:[invalid] style_emoticons/default/happy.gif)

My new server will have 2 Hard Disk Drives and I want to do a little more then just running 1 h@h.
In the end I want to have 3 Clients running, the autostart script mentioned above and also teh monitoring software G just G is using in the other thread.

Sooo... I guess the procedure will be about the same in the beginning. Update Ubuntu (Ubuntu 12 64 bit) then install Java.
Then I need 3 instances of h@h. My guess it is the best to have 1 h@h at one hd and 2 in the second.
Then we want to install the monitoring software and the startup script to the first hd.

Problem is I have absolutely no clue how to create such a script and how to add it.
Same for the monitoring software! halp!

It depends on the setup you have.
If you've got a router (not those you buy from the store)/ firewall you can setup an web net monitor and grab the information from there.
Otherwise ntop (advanced), iptraf or bmon will suite you fine.

For Ubuntu the standard suggest;
upstart.

For that it would be something like this:

CODE

#/etc/init/hath.conf

# Sets when it shall startup and shutdown.
start on startup      # You will need to check these are still used
stop on shutdown     # ------------------------||---------------------------

# Will restart the client incase daemon is killed.
respawn
respawn limit 99 5

script

user=' the user you want to run H@H '

exec su ${user} -c "/usr/bin/java -jar /home/${user}/HentaiAtHome.jar --silentstart"
end script

# This is will only execute after you've started the script
post-start script

end script

Set the upstart in the init conf file (I think..)

And not to sound too much like an asshole.
If you don't have the time for Linux, maybe you should have stay with Windows/Mac OS X..

But if you do got the time read this:
[upstart.ubuntu.com] http://upstart.ubuntu.com/cookbook/
or
[help.ubuntu.com] https://help.ubuntu.com/community/UbuntuBootupHowto
in case this does not work.

This post has been edited by GanGun : Jan 1 2013, 16:46

Hi all,

My H@H clients on Linux are running well but I just found out that they're eating up lots of RAM. I just wonder if this is normal.

On a Debian 6 one it simply takes up nearly all:


Similar for another CentOS 6 one:


But for another CentOS 6 one with double RAM, it just takes up half:


I am using Sun (Oracle) Java 1.6.0.38 with --disable_bwm --disable_logging options. Any advice?

OMG help me please!!!

I used this guide to set up 3 Clients on one Server except i only made 3.

CODE

so you'll want to create 5 users

command is: adduser <name> in the steps I'm sending you you'll want to use h1,h2,h3,h4,h5 etc

you'll want to place your H@H clients in the home folders of the users then navigate to each users folder and type. When you buy your xirvik server and add the ssd ask them to please mount the SSD as /ssd
"ln -s /ssd/cache1 cache"
then chance to the next H@H home folder (ie. h2) and do
"ln -s /ssd/cache2 cache"
each time you'll be incriminating the /ssd/cache number.

THEN you'll need to type the following after you're done with cache 1-5
chmod 7777 /ssd -R
this will set permission to allow any user to write on that drive.


then type this
iptables -t nat -A POSTROUTING -m owner --uid-owner h1 -j SNAT --to-source <ip address>
iptables -t nat -A POSTROUTING -m owner --uid-owner h2 -j SNAT --to-source <ip address>
iptables -t nat -A POSTROUTING -m owner --uid-owner h3 -j SNAT --to-source <ip address>
iptables -t nat -A POSTROUTING -m owner --uid-owner h4 -j SNAT --to-source <ip address>
iptables -t nat -A POSTROUTING -m owner --uid-owner h5 -j SNAT --to-source <ip address>

so that command tells anything run by the user (h1-h5) to run on <ip address> just change that to whatever you need. Then as the "root" user run the following commands..

open "screen"
once you've got screen open type screen 5 more times... guide each of them into the individual H@H folders for each one you want to run.

sudo -u h1 java -jar HentaiAtHome.jar --disable_logging --disable_bwm --max_connections=400
sudo -u h2 java -jar HentaiAtHome.jar --disable_logging --disable_bwm --max_connections=400
sudo -u h3 java -jar HentaiAtHome.jar --disable_logging --disable_bwm --max_connections=400
sudo -u h4 java -jar HentaiAtHome.jar --disable_logging --disable_bwm --max_connections=400
sudo -u h5 java -jar HentaiAtHome.jar --disable_logging --disable_bwm --max_connections=400


complicated yes but it's the only way to make it work

Worked well until this step:
chmod 7777 /ssd -R

It did not worked and gave me this error:

CODE

chmod: cannot access `/ssd': No such file or directory

Proceeded anyways and went up to this step:
sudo -u h1 java -jar HentaiAtHome.jar --disable_logging --disable_bwm --max_connections=400
where it asked me for Client ID and Key. Entered these and then this error popped up:

CODE

2013-01-04T17:04:32Z [ERROR] {org.hath.base.HentaiAtHomeClient.dieWithError(Hent
aiAtHomeClient.java:271)} Critical Error: Failed to initialize the cache.
2013-01-04T17:04:32Z [info] Shutting down...
2013-01-04T17:04:32Z [info] I don't hate you
root@serverIP:/home/h1#

Halp!
I think i need to give my users the rights to right on the drive?

This post has been edited by Ichy: Jan 4 2013, 19:08

/ssd is the folder where I have H@H installed so if your H@H client is in /home/h1 change it to chmod 7777 /home/h1 -R

If you notice in the steps...
"When you buy your xirvik server and add the ssd ask them to please mount the SSD as /ssd"

that would create a folder named ssd (remember / means root directory its a bit like c:\ on a windows machine so c:\windows would be /windows in linux **this statement is completely wrong but serves to illustrate an example**)

Ichy - type this df -h --- make sure you've got two drives in the G (Gigabyte) size. You'll want to use both hard drives for H@H so lets say df -h looks like this....

(IMG:[s1.directupload.net] http://s1.directupload.net/images/130104/v2r9dheu.png)
You'll want the first H@H client (calling it h1) on /main and client two (h2) on /ssd ... You should load up the 2nd disk (the one not handling / in my example it's /dev/sda1 *you can tell this from the first row*) with additional clients as whichever drive is running / will also handle journaling. If both your drives aren't there it's probably not mounted. If its not google mounting a new drive **Linux Distro** for help mounting your drive.


Mnching ignore it - Java sucks up TONS of virtual memory because it can but it's not actually allocated. -- notice how it says java is using 7.5% of your memory? that's what it's actually using. If some other program needed memory Java would give up memory from its "SHR" pool (shareable) .... Don't worry about it!

This post has been edited by G Just G: Jan 4 2013, 20:09

Noooooooo!

CODE

2013-01-04T18:00:35Z [info] CacheHandler: Loading cache.. (this could take a whi
le)
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)} java.lang.NullPointerException
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)}         at java.util.ComparableTimSort.sort(ComparableTimSort.ja
va:146)
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)}         at java.util.Arrays.sort(Arrays.java:472)
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)}         at org.hath.base.CacheHandler.populateInternalCacheTable
(CacheHandler.java:379)
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)}         at org.hath.base.CacheHandler.initializeCacheHandler(Cac
heHandler.java:242)
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)}         at org.hath.base.HentaiAtHomeClient.run(HentaiAtHomeClie
nt.java:114)
2013-01-04T18:00:35Z [ERR] {java.lang.Throwable$WrappedPrintStream.println(Throw
able.java:747)}         at java.lang.Thread.run(Thread.java:722)
2013-01-04T18:00:35Z [ERROR] {org.hath.base.HentaiAtHomeClient.dieWithError(Hent
aiAtHomeClient.java:271)} Critical Error: Failed to initialize the cache.
2013-01-04T18:00:35Z [info] Shutting down...
2013-01-04T18:00:35Z [info] I don't hate you

/edit: I already tried starting up the other clients with the second and third IP but the same thing is happening! (IMG:[invalid] style_emoticons/default/cry.gif)
/edit2: maybe this step has to be changed too? "ln -s /ssd/cache1 cache"

This post has been edited by Ichy: Jan 4 2013, 20:07

I need to see your setup.

type "df -h" and give me the output of that.

Goto /home and type ls -l and show me that.

This post has been edited by G Just G: Jan 4 2013, 20:17

df -h made this:

CODE

root@server88-208-193-162:/home# df -h
Filesystem                       Size  Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00  226G  1.5G  213G   1% /
udev                             1.8G  8.0K  1.8G   1% /dev
tmpfs                            721M  412K  721M   1% /run
none                             5.0M     0  5.0M   0% /run/lock
none                             1.8G     0  1.8G   0% /run/shm
/dev/sda1                        236M   33M  191M  15% /boot
root@server88-208-193-162:/home#

and ls -l


/edit:
root@server88-208-193-162:/home# ls -l
total 16
drwsrwsrwt 4 h1 h1 4096 Jan 4 16:54 h1
drwsrwsrwt 3 h2 h2 4096 Jan 4 17:51 h2
drwsrwsrwt 3 h3 h3 4096 Jan 4 17:53 h3
drwsrwsrwt 2 h1 h1 4096 Jan 3 09:30 installer
root@server88-208-193-162:/home#

This post has been edited by Ichy: Jan 4 2013, 20:21

Ichy - button top right area above "new thread" called Live Chat - find it and push it.

The Chat had an error now I Cant go back in (IMG:[invalid] style_emoticons/default/dry.gif)

CODE

19:25] == Disconnected from server: Connection to IRC server lost.

/edit2: maybe this step has to be changed too? "ln -s /ssd/cache1 cache"

You need to point the cache location to the real one? Doesn't seem like you need to create a symbolic link and points to somewhere else for this case as you have just 1 mount point "/". Try removing it (in "/home/h1" directory run "rm -rf cache" then run "mkdir cache", and perhaps need to run this as well "chown -R h1:h1 cache".) Then try running the client again?

This post has been edited by mnching: Jan 4 2013, 21:24

Already running thanks to G Just G! (IMG:[invalid] style_emoticons/default/smile.gif)

BTW G Just G I just wonder if you have a study on whether "--use_more_memory" would be useful?

BTW G Just G I just wonder if you have a study on whether "--use_more_memory" would be useful?

Preformance boost.
The H@H client will be able to allocate more memory for more users.

Preformance boost.
The H@H client will be able to allocate more memory for more users.

I know but just wondering how significant. I saw it used a little bit more memory but seems not significant in performance gain. G Just G also didn't put this in his tutorial.

Sometimes a seedbox manages to derp any upgrade process. It said something about overhead head or shit. Not sure since it happened so fast and it kept upgrading anyways. Not even sure if this is a problem.

But unlike other Boxes it refused commands like sudo apt-get install openjdk-7-jre or wget.
I managed to get it running using the knowledge in this thread but i'm asking, is there a way to fix it up so it works properly? (IMG:[invalid] style_emoticons/default/tongue.gif)

Sometimes a seedbox manages to derp any upgrade process. It said something about overhead head or shit. Not sure since it happened so fast and it kept upgrading anyways. Not even sure if this is a problem.

But unlike other Boxes it refused commands like sudo apt-get install openjdk-7-jre or wget.
I managed to get it running using the knowledge in this thread but i'm asking, is there a way to fix it up so it works properly? (IMG:[invalid] style_emoticons/default/tongue.gif)

apt-get is a program. If it doesn't have it you need to install it. Find out what OS you're running and Google "installing apt-get *" and insert your distro in to the asterisk.


Adding use more memory reduces disk I/O. Tenboro AFAIK hasn't explained how it does this just that it does. I run with it on but have noticed no difference.

Adding use more memory reduces disk I/O. Tenboro AFAIK hasn't explained how it does this just that it does. I run with it on but have noticed no difference.

I did see that it uses slightly more memory (Unix, originally something around 6% and becomes 8-9% after using it, clearly not as significant as I expected to be, seems that I can still use this option even I only have 128MB RAM...) but I didn't have time to study seriously if it really reduced I/O. My Unix knowledge is very limited (and I don't want to spend much time on that).

This post has been edited by mnching: Jan 11 2013, 20:07

apt-get is a program. If it doesn't have it you need to install it. Find out what OS you're running and Google "installing apt-get *" and insert your distro in to the asterisk.

oh I see. Thanks (IMG:[invalid] style_emoticons/default/happy.gif)

Also have anyone noticed this? Sometimes my h@h page says a server is on timeout! when I visit the screen page it just sits there doing nothing!
I have to restart it to make it work again.

This already happened 2 times in the past 3 month (IMG:[invalid] style_emoticons/default/cry.gif)

oh I see. Thanks (IMG:[invalid] style_emoticons/default/happy.gif)

Also have anyone noticed this? Sometimes my h@h page says a server is on timeout! when I visit the screen page it just sits there doing nothing!
I have to restart it to make it work again.

This already happened 2 times in the past 3 month (IMG:[invalid] style_emoticons/default/cry.gif)

I happened more times. Sometimes it is just some network interruption and if the process is still running, it should get back to "Online" eventually. If you don't want to restart and you want it recovered immediately, then kill the process and re-run the startup script.

(1) Look for the process number using command "ps -ef | grep java", you should see something like this.



The process number is 29708 here.

(2) Issue command "kill 29708".

(3) Check again using command "ps -ef | grep java" until the process disappears, then use your startup script to start it again (e.g. "java -jar HentaiAtHome.jar").

Not sure about your configuration but if you put that on startup already the simplest way could be just restart the server.

The timeout issue happens to me all the time. Ignore it - it fixes itself. If you want to just restart your H@H server. It's a bug and Tenboro is aware of it.


https://forums.e-hentai.org/index.php?showtopic=91503

This post has been edited by G Just G: Jan 13 2013, 10:59