Errors, bugs, questions - page 2602

 
Roman:

Apparently MT4 has also been switched to the new Clang compiler, in MT5 the latest beta build is said to also load the system.
If you are handy with a soldering iron, replacing the correct capacitors should be no problem.

They are, but not very much. And without knowing which capacitors have to be replaced, it's not a feasible idea. Plus, you have to go farther for the capacitors than for the power supply...

 
Alexey Viktorov:

ps; Then it turns out that build 1220 puts more strain on the power supply than build 1190? If so, the developers would do well to pay attention to this.

It is impossible for a CPU to drop the 12V power supply line

If PSU is faulty, it does not deliver all the ratings (12.5, 3.3), then there will be failures in memory, then the CPU, and only then the hard drives, fans.... If you have a video card, it may start to fail if the 12 V sag, but the first sag at 12 V will feel the SSD, and other devices work without problems ... imho, it doesn't happen that way...although.... thought it was the hard drive when the memory failed, even win7 couldn't install to the hard drive under suspicion, at 30% of the installation it wrote can't copy the file... the problem was caused by faulty memory

ZS: wrote about AIDA64 - did you do it?

 
Igor Makanu:

ZS: wrote about AIDA64 - did you do it?

Can't remember what it's about? Didn't do it then.

psps; Didn't see the additions to the text.
 
Alexey Viktorov:

I don't remember what this is about? Didn't do it then.

psps; Didn't see the additions to the text.

download from the off. site, there trial version with limited functionality, but stress tests are available, first test without hard drives, then add the checkbox to test the hard drive, there in the log AIDA64 all writes

SZZY: if you had during power down writing to SSD maybe and that was the problem for "SSD crash" - invalid data in the sector, the drive controller can try to fix when reading / writing it, and can freeze - would have long ago chkdsk performed, maybe the problem and would go away .... but already with a soldering iron to fix the PC ))))

 
Igor Makanu:

download from the off. site, there trial version with limited functionality, but stress tests are available, first test without hard drives, then add the checkbox to test the hard drive, there in the log AIDA64 all writes

SZZY: if you had during power down writing to SSD maybe and that was the problem for "SSD crash" - invalid data in the sector, the drive controller can try to fix when reading / writing it, and can freeze - would have long ago chkdsk performed, maybe the problem and would go away .... but already with a soldering iron are going to fix the PC ))))

Downloading. I will run a test now, but how do you explain the fact that build 1190 has been running for almost 5 hours without any problems and this morning when I ran build 1220 it kicked the drive out of system instantly? I don't need to tell you about the repeated disk crash yesterday.

I'll run the test again later on build 1220 and see what happens.

 
Alexey Viktorov:

Downloading. I am going to check it now, but how do you explain the fact that build 1190 has been running for almost 5 hours without any problems and the drive crashed instantly this morning while running build 1220? The repeated kicking out yesterday is unnecessary to say the least.

I'll run the test on build 1220 again later and see what happens.

I'm not a magician, but I can say again - probably write errors in the SSD and files from one build are not affected, but the other is not correctly written

 
Igor Makanu:

I'm no wizard, but I can say it again - there may be writing errors in the SSD and files from one build are unaffected and the other is written incorrectly

So I compile this file periodically, run the test, edit, compile and so on... And in the terminal of build 1990 the file was copied that was tested in 1220.

 
Alexey Viktorov:

So I periodically compile this file, run the test, edit, compile and so on... And in the terminal build 1990 file was copied that was tested in 1220.

not an indicator at all, files are cached both by OS and controller when reading, caching algorithms can be different, both file layout tables and small files can be cached

If you want to check operability of 1220 build unambiguously - install new terminal on C-drive and test it there, then questions about hardware operability will disappear.

 
Igor Makanu:

not an indicator at all, files are cached both by OS and controller when reading, caching algorithms can be different, both file layout tables and small files can be cached

If you want to check 1220 build unambiguously - install new terminal on C-disk and test it there, then questions about hardware operability will disappear.

"We'll go another way." I've ordered a UPS and a power supply, I'll replace it and then I'll think about it. If there was a problem turning the computer on after a blackout, then there must be something wrong with it. And why should I wait for it to die completely?

 

Greetings Gentlemen, here is a question:

I start MT5terminal from the command line, with a configuration file, immediately start the tester, after the pass the tester saves the report, and closes the terminal, in the folder with the report appears another file debug.log with the error "[1110/121145.068:ERROR:crash_report_database_win.cc(428)] unexpected header", what is this error?

Files:
debug.log  1 kb
Reason: