| Author |
Thread Statistics | Show CCP posts - 18 post(s) |
|
CCP Sharkbait
   |
Posted - 2007.07.11 14:46:00 -
[1]
i need more logfiles of this happening please.
when you are running logserver to log this, please go to tools and enable serverModeSave. then set this to 100k entries. this will stop your machine lagging while you are playing. please get all the files together and attach them to a bug report with the title "desync issue" and put a good description of what you was doing in to the description field.
thx
/Shark
|
|
|
CCP Sharkbait
|
Posted - 2007.07.12 15:54:00 -
[2]
 Originally by: Venkul Mul
Edited by: Venkul Mul on 12/07/2007 08:16:38
Originally by: CCP Sharkbait
i need more logfiles of this happening please.
when you are running logserver to log this, please go to tools and enable serverModeSave. then set this to 100k entries. this will stop your machine lagging while you are playing. please get all the files together and attach them to a bug report with the title "desync issue" and put a good description of what you was doing in to the description field.
thx
/Shark
Will try to get them.
Sometime it is not easy to detect when the desyng started (for example if you are probing a system or mining). So how far back we should go with the logs?
I mean I notice the desync at 20.10 hours, my logs have been done at 20.08, 20.02, 19.55 (only an example, I don't know the effective time needed to cover 100K entries), I should send only the current and precedent log (20.08), the last 10 minutes or more?
Edit: please fix the forummtoo. 3 attempts to get this post in.
basically i just need a logfile of the desync starting and then happening. not too bothered about when it finishes, i just want to know what causes it more than anything.
|
|
|
CCP Sharkbait
|
Posted - 2007.07.13 20:23:00 -
[3]
logserver is old now and it's due to be retired. we are working on a new logserver util. not sure the state of it or when it can be released yet.
|
|
|
CCP Sharkbait
|
Posted - 2007.07.15 01:38:00 -
[4]
please keep sending in the logfiles, we still need all the information we can get.
|
|
|
CCP Sharkbait
|
Posted - 2007.07.16 20:58:00 -
[5]
ok, i will give you all the bad news. i have ran out of beer at home.
other news the fix didn't work, but that not bad news as such. we know where the problem is now and there is a lot of focus going into fixing it. once we have a fix and QA feel it's safe to deploy, the we plan to get it out onto the server asap.
once i have more details, i will spam the forums with them. if you want to keep submitting bug reports with logfiles, please do so. they will not be deleted. they might even help further in the matter.
|
|
|
CCP Sharkbait
|
Posted - 2007.07.16 21:34:00 -
[6]
it's not me fixing it, Richard is handling it. i feel safe to bet he will have it fixed tomorrow. if anyone can fix it, it's him.
there is a wee problem where this requires a client patch, it's too late to get it out in tomorrow's patch now. but there is a patch for the python2.5 upgrade and the target date for that is a week tomorrow.
once i have more details, you will also have them
|
|
|
CCP Sharkbait
|
Posted - 2007.07.17 16:50:00 -
[7]
there has been another fix for the issue checked in now from Richard. but it's a big change and requires alot of testing. once i know more i will inform you
|
|
|
CCP Sharkbait
|
Posted - 2007.07.17 19:26:00 -
[8]
Originally by: Pardack
Edited by: Pardack on 17/07/2007 17:08:45
Originally by: CCP Sharkbait
...there is a patch for the python2.5 upgrade and the target date for that is a week tomorrow.
Are you doing a test to compare python 2.5.1 to the older (2.1?) version? The last test was cancelled and I haven't seen a new one announced, did I miss it? If so, I'd love to see some before/after numbers from the tests. If not, is the test going to happen? I was under the impression you wanted to load test and get an accurate before/after comparison to make sure it was actually helping and not hurting performance.
we have been planning these, but both code and network problems have stopped us. i was doing some benchmarking on a localserver and now i have finished with some of the main things i wanted to do, i will be continuing tomorrow and hopefully have some numbers to publish at the end of the week. there will also be tests again on sisi now trying to benchmark between the language versions. but from what we have seen so far, there is a reducion in CPU and memory usage when using python 2.5.1 against python 2.3.3.
another quick note, the 1st tests for the Desync issue was a success and it appears the fix has worked. there will be more tests for this throughout the week.
|
|
| |
|