Am I alone in thinking that the Java Quick Starter is a total CPU killer? When it’s on, my computer struggles to play video or even an mp3 file! I had it turned off, however, after the last update (1.6.0_13), it turned back on! No Sun, I turned it off for a reason!
The only blessing is that the Java Quick Starter is only for WinXP and Win2000.
BTW: To turn it off, go to Control Panel -> Java -> Advanced Tab -> Miscellaneous -> Untick “Java Quick Starter”
…The java updater too; disabling it through the control panel interface does not remove the process from launching at windows start-up - just makes it do nothing when launched.
Then there is Webstart; pollutes my start menu, desktop & uninstall list without even asking me!
Several of which have become unremovable; presumably due to a Java update in the past 12 months that has broken the webstart cache somehow.
Java is definitely not the most well behaved of applications… but atleast the installer doesn’t try to bundle lots of other unrelated or useless apps along with it, as is becoming the case for many others.
well not to worry, quick starter is no longer needed in windows vista+, as soon as xp dies so will quick starter.
try ang get a program called add remove pro it gets rid of broken links failing that run regedit(if your using windows).
Seriously add rem pro is good it gets rid of orphaned reg files too if you tell it otherwise it acts just like windows addremove but with the addition of remove broken link.
I’m not on windows anymore but I’m pretty sure there was a program provided with windows or free for download from ms to do just that. in never versions of windows it checks automatically for it’s self.
My WinXP box just upgraded to Java 1.6.0_15. Music and videos again stopped playing smoothly, CPU usage kept spiking. I went into into Java, yep, friggin Java Quick Starter is back on. Hello Sun!?! Do you even test this stuff before putting it out?! Turned it off. CPU stopped spiking and mp3s and video play no problems.
Will only a problem for people running WinXP.
I hope you understand that we don’t see this behavior. Of course this stuff is tested. (for example, my workstation is an XP system, and I have never experienced anything like this). Somehow your particular configuration triggers this problem. AFAIK quick starter is supposed to turn itself off if it notices high system load. Not that it’s supposed to be doing much anyway, it just touches a few files on the hdd so that they’re in the file cache.
It would be more constructive if you thought about what could be making your system different from a typical system. Any specific sw installed? What’s your hw configuration (cpu/memory). Any other resident java processes? Any other ideas?
Hi Trembovetski,
I Googled “Java Quick Starter cpu xp” and saw that a lot of people have this problem. I also just googled “Java Quick Starter” and saw that it also mostly came up with people having trouble with it and wanting to know how to get rid of it.
Maybe there is something odd about that computer. No clue what it would be, it’s a dual core laptop that I leave on 24/7 as a webserver, skype client, and I play music and video from it. Only s/w on it is Tomcat, Skype, Winamp, VLC, VNC, and Cannons remote camera operation s/w.
My post was mostly out of frustration, as when software upgrades itself, it should keep the users preferences (ie, not turn JQS back on if the user turned it off).
On a side note. The whole concept of forcing a program to stay in memory IMO seems counter productive. If the O/S swaps Java out of memory, it means the person isn’t using it much, Java shouldn’t force itself back into memory on the chance a user may want to do something with it.
Anyway, rant over. I do think Java is awesome, and these little things aren’t really all that bad. Would be nice if they were fixed, that’s all.
Cheers.
[quote]Several of which have become unremovable; presumably due to a Java update in the past 12 months that has broken the webstart cache somehow.
[/quote]
I suggest trying CCleaner to remove those uninstalls, it has a option to remove broken uninstalls.
The problem being that Windows does an extremely crappy job at this sort of thing and basically will force nearly everything out of RAM so you get lots of swapping. It’s like someone at Microsoft didn’t get the memo that RAM was faster than DISK and implemented their virtual memory system backwards. (Actually I think they are just way too aggressive at using RAM for a filesystem cache - caching all the files you aren’t accessing and forcing your software to the paging file so you visit swap central every time you switch apps.)
Haha. Yes, I totally agree. However, IMO, I think the JQS is not a good solution. I would prefer if Java just lived with the Microsoft crappyness (or harassed Microsoft to fix it).
Like… seriously… somebody notifies you of a serious and blatantly annoying bug and this is how you respond? Way to go! What about diving into the code straight away?
Exactly how do you expect me to respond to a post that something is wrong without the details about the environment and the conditions to reproduce the bug? Dive into code and look for what? enableEvilCPUConsumption() function? (Not to mention that I don’t work on jqs even remotely.)
I don’t know about your approach to fixing bugs, but I have already stated that this one can’t be reproduced in house so we need more info in order to understand what could be wrong.
Anyway, people familiar with the matters told me that there’s a logging mode in the jqs that could be enabled which could help with figuring out what’s wrong. I’ll get back with the specifics, although I believe you could just run jqs.exe /help (I can’t verify right now, I’m on mac).
jqs is supposed to back off if there’s cpu or memory pressure in the system.
Dmitri
Well, you know how it is with the internet. You won’t find posts about how awesome jqs is because those people don’t bother posting =) Surely, among millions of java users there would be a small percentage of people running into issues. We don’t see a flood of bugs on this, so it is likely that the percentage is small. We’ve heard about a few cases, mostly in corporate environment (so it may be related to privileges or something).
OK, thanks for the details. What about your user privileges? Does your account have administrator privileges? There’s a suspicion that jqs may not know that the system is busy because the processes taking cpu/memory are out of its scope somehow.
The best way would be if you could enable the logging mode and try again (I’ll get back to you on how to do that).
Dmitri
Could you please do this:
jqs.exe -unregister
jqs.exe -register -verbose 3 -logfile ${JQS_LOGFILE}
jqs.exe -enable
(where JQS_LOGFILE is a path to a file) and see if you could reproduce the problem.
Very true.
Yes, I have admin privilege. I should also point out that Task Manager does not show the CPU going crazy when JQS is running. It shows a spike of 1 or 2% every few seconds (while with JQS off, it pretty much stays a 0 or 1%). The fact that it interferes with WinAmp and VLC is a bit of a mystery.
Honestly, I don’t mind just turning it off. I never need it, the computer is just a webserver/skype/media player. However, the last 3 Java updates have all turning it back on by default, and every time it turns it back on, I have spent a 10 minutes trying to work out why the computer isn’t playing my media properly any more (short term memory :)).
Anyway, I’ll be happy to run any logging that you would like me to.
Cheers.
EDIT: Running tests now.
Ok. Done. Played 1 song in Winamp. Song skipped 3 times (2 times were only small skips around 0.2 seconds, but 1 skip was probably 1 second long).
Log file attached.
Quoted for truthery. I hate having to run jcontrol to allocate more memory to applets every time I update Java.
Thanks for the log, I’ll forward it to the team. This is really bizarre that it affects media that much. What kind of hard drive do you have?