- This topic is empty.
September 29, 2016 at 4:45 pm #54269
This time is good.
200.76 Kbit/s (Connections: 1007 Load: 8%)
CAN I overwrite bitrate with this file _dbg???
_dbg will not work, UI expect one line, this file will output more. The only difference between this files are 3 print commands that should not do anything about the final result.
Also now load is low, we should redo it when load is going over 100.
You can comment the print statements (add # in front of the print like this #print ) one by one to see when is starting to behave wrong again.
To see the differences you can use diff command
diff bitrate bitrate_dbg
> print "Load Line=@Load", "n";
> print "CPUs=$CPUs","n";
> print "Load=$LOAD","n";
September 29, 2016 at 5:52 pm #54270
Ohh, you have all right. Today I restarted zeroshell and the load is little. Normally it grow up in time. Tomorrow I will do the test again.October 2, 2016 at 7:06 am #54271
Dear IulyB, actually, after 30 approx running hours, I don’t have more impossible load % (I’m at 6-10% with 2354 session). I didn’t activated your script in scripts/chron so I don’t think it is working….
What can be happened?
I changed ftp server solution in these days and as I told you I’m uploading 1million of little files a day with continuous multisession ftp opened. I worked in the last time also on a voip PBX in line with 20 users. Can be possible that the wrong load in zeroshell has been created from ftp upload server or voip not good configured sessions?
We will see in the next days… I don’t know what think… . I’m sorry that a critical production exigence obbligated me to change the reference situation before to understand the problem.October 2, 2016 at 3:51 pm #54272
You had 2 distinct issues, you had a big real load around 100 that I can’t replicate easy on my APU, my old Alix however had loads of 100 very frequent when I had multiple torrents but I never check them against command line.
The second issue was that your load was displayed wrong on ZS interface.
Your issue may also be due to some interaction with other package that was installed on ZS.October 5, 2016 at 1:23 am #54273
I just noticed a related bug. Te interface started to show ERROR. However the bitrate script returned right.
34.69 Mbit/s (Connections: 1183 Load: 7%)
I believe this is from kerbinet script. Unfortunately I don’t have the code for that script.October 5, 2016 at 5:36 am #54274
Dear Mr. Iulyb, after 4 days of Zeroshell run I don’t have impossible load more. Actually I have 8,5Mb/s, 2361 connections, 6% load. I’m so sorry that I cannot more reproduce the problem. In this time I changed some ftp test server solutions and voip pbx setting. I don’t know if the problem can arrive from something not perfect on their config… I’m so sorry, if I can help you….
Here the picture of the problem.
October 5, 2016 at 2:56 pm #54275
Are having 1G of memory? Then why do you use a 64 kernel?
Unless the application is optimized against memory, a 64 will consume more memory that a 32 bits will consume just because of addressing.
Also making the kernel 64 and keeping the rest of apps at 32 in ZS does not make a 64 distro.October 24, 2016 at 4:09 pm #54276
Unfortunately never had again this iussue. My conclusion is that I had an abnormality in my network/software that probably created zeroshell to fault. Zeroshell worked all time fine and that load was not real I think.November 4, 2016 at 4:25 pm #54277
Hallo IulyB I have the error again!!!
184.38 Kbit/s 141 connections Load 2504%
Do you need some tests?November 4, 2016 at 4:33 pm #54278
Re-read the thread again, to get an idea what you did and what u got then run from command line that debug script ( terminated in _dbg) I did and post the results.
BTW, is interesting that u have 2503, 2504 .. how about 30xx or 20xx ? Did u see this values?
- You must be logged in to reply to this topic.