Since The WABBIT noticed it to, I thought I'd bring up the issue here in case you can provide any insight.

If TWGS is installed on a machine, and a telnet client using the Java Telnet Application (http://www.javassh.org), examples of which include J-TWAT and my Kokua, running on the same machine connects to the local TWGS, TWGS CPU usage shoots up to 90+% after telnet negotiation and at login, 100%.

There could be a bug in JTA where it performs improper negotiation, but I don't imagine TWGS should react in such a harsh manner. You can either test this by downloading the JTA client or by using Kokua - http://www.twdata.org/kokua and connecting to a local TWGS instance. Interestingly, if you also have TWXProxy on that machine and have the JTA client connect to it first, this issue doesn't come up.

Thanks for the help,

Don

quote:Originally posted by mrdon

Since The WABBIT noticed it to, I thought I'd bring up the issue here in case you can provide any insight.

If TWGS is installed on a machine, and a telnet client using the Java Telnet Application (http://www.javassh.org), examples of which include J-TWAT and my Kokua, running on the same machine connects to the local TWGS, TWGS CPU usage shoots up to 90+% after telnet negotiation and at login, 100%.

There could be a bug in JTA where it performs improper negotiation, but I don't imagine TWGS should react in such a harsh manner. You can either test this by downloading the JTA client or by using Kokua - http://www.twdata.org/kokua and connecting to a local TWGS instance. Interestingly, if you also have TWXProxy on that machine and have the JTA client connect to it first, this issue doesn't come up.

Thanks for the help,

Don


Well, I tried your test both ways using J-Twat. Both meaning with TWX and without, connecting to my own server using the same machine. TWGS & TW2002 CPU times didn't shoot up, but J-Twat was running anywhere from 39% to 59% all the time, regardless of TWX or not. I'll give it a test with the other one you mentioned and let you know what I see...

Celery 366x2 processors (yup, she's a dually)
512Meg PC133

Since I read that, and went looking at Kokua, I went to the bot page. Found something interesting. Exactly what you said here, but it was on the bot page not the main Kokua page.
http://www.twdata.org/kokua-bot/
Scroll all the way to the bottom...
Maybe there is something in the way they utilize the telnet session in Java? Donno...

Well, as the author of both Kokua and Kokua-bot, that could be the case I thought I remember the problem occuring with J-TWAT, but perhaps I was mistaken. Wabbit mentioned this problem too, but I'm not sure how he was using JTA.

quote:Originally posted by mrdon

Well, as the author of both Kokua and Kokua-bot, that could be the case I thought I remember the problem occuring with J-TWAT, but perhaps I was mistaken. Wabbit mentioned this problem too, but I'm not sure how he was using JTA.


Well, I tried this also with Kokua, but not Kokua-bot. I got the same result as when I was using J-Twat. No significant increase in TWGS or TW2002 cpu usage. However, the Javaw process would, at times, reach peaks of 70%.

Anyways... Just looked at the projected path for Ivan... I'm right in the middle of the way... Gotta go lock up what I can and move the rest inside... That means no more TWGS server until this thing clears. SORRY GUYS... I might not have a house when this thing gets done, but I'm taking my computer with me if I have to go anywhere. Wish me GOOD LUCK (not just any luck will do...) Geeeesh, hurricanes...

Strange, you would certainly notice it as your system would grind to a _very_ slow crawl. My system:
- Java 1.4.2_04
- Windows 2k, all patched
- Latest TWGS

Don

quote:Originally posted by mrdon

Strange, you would certainly notice it as your system would grind to a _very_ slow crawl. My system:
- Java 1.4.2_04
- Windows 2k, all patched
- Latest TWGS

Don


Very strange. Mine didn't slow down at all...
- Java 1.4.2_05
- Win XP Pro (all patches EXCEPT SP2!)
- TWGS 1.01.55 (TW2002 3.13 - Latest)
~special note, but then again, it might be the dual processors?~

quote:Originally posted by mrdon

Well, as the author of both Kokua and Kokua-bot, that could be the case I thought I remember the problem occuring with J-TWAT, but perhaps I was mistaken. Wabbit mentioned this problem too, but I'm not sure how he was using JTA.


To answer your questions, I'm using JTA telnet wrapper for the connections to TWGS. No, I'm not using a display window. As the application will be fully automatic. So, there is no need for a display.

I have not modified any of the code in JTA. And I'm using the latest released version of JTA from http://www.javassh.org, which I believe is v2.5.

As I stated before, if the application and TWGS are on the same machine. TWGS pegs out the CPU, not the application. And then seems to freeze at the login prompt. I say that, as nothing sent to the server. Is every received or responded to, by the server. If I already have the TWGS window open. Then I must stop the server, before the system returns to normal. But, it takes 5-15 minutes before the server is stopped.

Yet, if I do not have the TWGS window open. I must shut down TWGS from the Task Manager. As it takes over 20 minutes before I can get the TWGS window to open.

I have 3 systems, 2 of which I have tried it on. And in both cases, if the application w/JTA and TWGS are on the same machine. I have this problem. But, if they are on seperate machines. Communicating between Lan or Internet. Then the application is able to connect and login into the server, with out any problems.

Specs of systems used -

System 1:

Windows 2000 AS w/ SP4 and rest of security fixes
AMD Athlon Duron 1Ghz CPU
1.5GB Ram
>400GB HD space, between 6 HDDs with over 70% of total space is available.


System 2:

Windows XP Home Edtion, OEM w/ SP1
AMD Athlon XP 2.2Ghz CPU
512MG Ram
60Gb HDD, with 85% available

Both system have totally different software installed. And have totally different processes running in the background, with the exception of the required Windows processes.