[SOLVED] CT doesn't start anymore after changing Java's RAM size Thread poster: MikeTrans
|
|---|
MikeTrans
Germany
Local time: 15:51
Italian to German
+ ...
Hi,
I'm quite impressed after running a project like described in the help section (Quickstart). I have used a single little TM of about 10.000 entries and a Memory for terms. I was able to finish the project quickly and translating about 30 segments of a simple localization text.
For a second project, I decided to use the following setup for trying out the program:
- Importing into a Glossary about 9000 tab separated entries, read-only
- Memory for ... See more Hi,
I'm quite impressed after running a project like described in the help section (Quickstart). I have used a single little TM of about 10.000 entries and a Memory for terms. I was able to finish the project quickly and translating about 30 segments of a simple localization text.
For a second project, I decided to use the following setup for trying out the program:
- Importing into a Glossary about 9000 tab separated entries, read-only
- Memory for terms of the last project, 20 segments
- A small Memory for segments, 10.000 segments
- Large memory, memory for segments, pretranslation mode, read-only, 333.000 segments
I've loaded the big memory at the end. First, 140.000 segments are imported fast (seeing that on the progress bar), but CT remains stucked at 143000. My task manager shows that only 1,3 GB of physical memory is used.
I had to kill the program with the task manager.
After cleaning all system Temp directories and restarting, I now go into Edit > Options and change the first entry, Ram size to 2048.
CT will not start anymore.
My system:
Laptop with Windows 7 Home, i5 Intel Core, 4GB RAM
Note:
I don't think it's important, but after the 1st project, I have changed the Fuzzyness for memories from 33 to 25,
and the Autocomplete settings from 3-8 to 2-4.
Thanks for helping,
Greets,
Mike
[Edited at 2012-06-18 14:01 GMT]
[Edited at 2012-06-18 23:56 GMT] ▲ Collapse
| | | | Igor Kmitowski 
Poland
Local time: 15:51
Member (2016)
English to Polish
+ ...
| 32-bit vs 64-bit Java | Jun 18, 2012 |
Hello,
The 32-bit version of Java Runtime has about 2GB RAM memory limit so if you need more RAM to load large Translation Memories you should download and install 64-bit Java Runtime for your system to overcome this limitation. Of course, your operating system should be 64-bit as well.
If you wish to keep running CafeTran on the 32-bit Java Runtime please do the following:
1. Go the program's installation folder and click Cafetran.jar file. This mode takes... See more Hello,
The 32-bit version of Java Runtime has about 2GB RAM memory limit so if you need more RAM to load large Translation Memories you should download and install 64-bit Java Runtime for your system to overcome this limitation. Of course, your operating system should be 64-bit as well.
If you wish to keep running CafeTran on the 32-bit Java Runtime please do the following:
1. Go the program's installation folder and click Cafetran.jar file. This mode takes only 64 MB of RAM and lets you change the settings in Edit | Options. Next, decrease Java Memory Size value in Options below 2GB until the program can start normally.
I hope it will help.
Igor
MikeTrans wrote:
Hi,
I'm quite impressed after running a project like described in the help section (Quickstart). I have used a single little TM of about 10.000 entries and a Memory for terms. I was able to finish the project quickly and translating about 30 segments of a simple localization text.
For a second project, I decided to use the following setup for trying out the program:
- Importing into a Glossary about 9000 tab separated entries, read-only
- Memory for terms of the last project, 20 segments
- A small Memory for segments, 10.000 segments
- Large memory, memory for segments, pretranslation mode, read-only, 333.000 segments
I've loaded the big memory at the end. First, 140.000 segments are imported fast (seeing that on the progress bar), but CT remains stucked at 143000. My task manager shows that only 1,3 GB of physical memory is used.
I had to kill the program with the task manager.
After cleaning all system Temp directories and restarting, I now go into Edit > Options and change the first entry, Ram size to 2048.
CT will not start anymore.
My system:
Laptop with Windows 7 Home, i5 Intel Core, 4GB RAM
Note:
I don't think it's important, but after the 1st project, I have changed the Fuzzyness for memories from 33 to 25,
and the Autocomplete settings from 3-8 to 2-4.
Thanks for helping,
Greets,
Mike
[Edited at 2012-06-18 14:01 GMT] ▲ Collapse
| | | | Selcuk Akyuz
Türkiye
Local time: 16:51
English to Turkish
+ ...
http://proz.com/forum/cafetran_support/216263-additional_fields_in_a_glossary.html#1875092
Igor Kmitowski wrote:
CT may freeze when you load huge TMX or glossary files into RAM memory assigned to CafeTran. You can increase this value. See Edit | Options | Memory tab | Java memory size (MB). Remember not to go over the actual RAM memory on your system.
Hi Mike,
Until Igor steps in and finds a solution, you can change the setting at Registry Editor.
My Computer\HKEY_CURRENT_USER\Software\JavaSoft\Prefs\translator
Change the setting at Java/Memory
Selcuk
| | | | MikeTrans
Germany
Local time: 15:51
Italian to German
+ ...
TOPIC STARTER | I was able to re-start, but... | Jun 18, 2012 |
Hi Selcuk,
I did this, but with no avail.
By deleting the Cafetran.conf file, the program will restart.
Also, by changing the 2nd value in the Cafetran.bat file, -Xmx2048m, and double-clicking the bat, it will also start.
If I go in the Edtit > Options > Memory, I see 2048 for the RAM value.
The question is: will CafeTran now behave as intended?
I'll quickly find it out by reloading my project.
Mike
| | |
|
|
|
Selcuk Akyuz
Türkiye
Local time: 16:51
English to Turkish
+ ...
Always launch the program through Start.jar. as advised by Igor.
In my system it is located at C:\Program Files\CafeTran Espresso\cafetran
You can create a shortcut for it.
| | | | MikeTrans
Germany
Local time: 15:51
Italian to German
+ ...
TOPIC STARTER | no big TM load | Jun 18, 2012 |
Ok, i have splitted the big thing into 3 x 100.000+ segments putting them into a single folder.
Opening the memory chosing the folder will again put CT to sleep.
Opening in sequence the 3 files will result in a fast load of part 1, but again send CT to sleep on part 2.
150.000 segments seem *at the moment* to be the limit with a 4GB Ram computer system.
Java may have some severe limits for program structures like CT. Igor, have you ever thought about ... See more Ok, i have splitted the big thing into 3 x 100.000+ segments putting them into a single folder.
Opening the memory chosing the folder will again put CT to sleep.
Opening in sequence the 3 files will result in a fast load of part 1, but again send CT to sleep on part 2.
150.000 segments seem *at the moment* to be the limit with a 4GB Ram computer system.
Java may have some severe limits for program structures like CT. Igor, have you ever thought about SQL databases or else? Not being a developing expert, I know that there are engines much faster than Java.
Greets,
Mike ▲ Collapse
| | | | Selcuk Akyuz
Türkiye
Local time: 16:51
English to Turkish
+ ...
| large TM loads here | Jun 18, 2012 |
Mike,
My system is rather old, a 6-year-old laptop with 2GB RAM only. I know I have to buy a new one but no time for installing all those programs and making the settings.
Java size is 1048 MB and I can easily use a TM with 300k segments. Actually, one TM with 300K, another with 80K and a termbase with 20k entries. And many other resource hungry applications are open in other windows.
And to make things difficult that large TM is not read-only!
| | | | MikeTrans
Germany
Local time: 15:51
Italian to German
+ ...
TOPIC STARTER | Why I'm using so "large" TMs... | Jun 18, 2012 |
Hi Igor,
thanks for responding. I didn't see your response until now, so I wasn't aware of these tips.
In fact, I have a 64 bit windows system, but I cannot tell if my Java 1.6.0 installed is also 64 bits.
I will download a 64 bit version of Java as soon as possible.
About large TMs:
I mainly use large TMs for concordancing purposes. Presently I'm using another CAT tool and I run a thrid-party program in the background which has the large TMs l... See more Hi Igor,
thanks for responding. I didn't see your response until now, so I wasn't aware of these tips.
In fact, I have a 64 bit windows system, but I cannot tell if my Java 1.6.0 installed is also 64 bits.
I will download a 64 bit version of Java as soon as possible.
About large TMs:
I mainly use large TMs for concordancing purposes. Presently I'm using another CAT tool and I run a thrid-party program in the background which has the large TMs loaded and where I can eventually go and see segments in context.
These large TMs are useless for showing me only fuzzy matches and would also considerably slowdown ALL of the standard CAT tools like Trados Studio, Déjà Vu, MemoQ, Trados 2007, SDLX.
I can confirm that, as I've been using all those programs, yet not being very satisfied by the amount of information shown from the TMs.
During my translations, usually 2-3 sentences out of 50 have a fuzzy match of 50% or more, so I'm really not getting excited about the Fuzzy sentence feature. Instead, good sub-segment information could tell me if it's worth searching the databases, or if I need to just start a more complex research online or in my specialized litterature.
As a curious example:
In order to make my current leading CAT tool (which I will not name) tell me about all sub-segments available in order for me not losing time searching them manually in the databases, I had to split up a large TM (330.000 segments) up in 6 parts alphabetically. These parts sum up to 5.600.000 short segments up to 5 words. All this, just to show me what's in the TM and - more importantly: what is *not*.
I think CafeTran actually does much better than this.
CafeTran Expresso 2012 goes far beyond simple Fuzzy Matching, showing sub-segments for which the programs shows context in an easy accessible fashion. And for me, this makes a huge difference!
THIS IS WHY it is so important that CT should also handle large TMs: with a fast engine, CT would be way ahead of the competition, which only introduces new features at a slow pace and ridiculous conditions and prices.
[ADDED]
---------------
Reading the post of Hans, it now becomes clear that I must have a serious system quirk involving my Java version (probably not 64-bit as it should be)
---------------
Thanks,
Mike
[Edited at 2012-06-18 22:40 GMT] ▲ Collapse
| | |
|
|
|
MikeTrans
Germany
Local time: 15:51
Italian to German
+ ...
TOPIC STARTER
[SOLVED]
For anyone with this trouble, having
Windows 7 Home, 64 bits
At least 4 GB RAM
a)
Install the latest Java JRE SE Runtime, taking care to download the 64 bit version.
http://www.oracle.com/technetwork/java/javase/downloads/jre7-downloads-1637588.html
b)
Within Caf... See more [SOLVED]
For anyone with this trouble, having
Windows 7 Home, 64 bits
At least 4 GB RAM
a)
Install the latest Java JRE SE Runtime, taking care to download the 64 bit version.
http://www.oracle.com/technetwork/java/javase/downloads/jre7-downloads-1637588.html
b)
Within CafeTran, change
Edit > Options > Memory > Java RAM size 1024 TO 2048.
Quit CafeTran and restart.
Large DBs will now load without delay. My 330.000 segments were loaded in 15 seconds.
Very happy to see that,
Mike
Note:
Consider not to remove other Java versions, but deactivate them in the Java panel.
Some other CAT tools still may still have to use much older 32-bit Java versions, especially Trados 2007, Trados Studio (Multiterm, Multiterm Extract). So, I cannot really advice you to remove them from your system for the advantage of the newest versions, as these tools will most probably NOT work as intended anymore if you do so.
[Edited at 2012-06-19 00:12 GMT] ▲ Collapse
| | | | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » [SOLVED] CT doesn't start anymore after changing Java's RAM size | CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
| | Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
