I keep getting X errors:
Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
Neal Becker wrote:
I keep getting X errors:
Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
I'm starting to suspect that google chrome is the culprit.
sudo lsof -c X | grep socket | wc
219 1971 16644
Now close chrome:
[nbecker@nbecker1 jupiter]$ sudo lsof -c X | grep socket | wc
60 540 4560
I'm having no such problem with Chrome. However, I have noticed, that on occaission, Chrome being Beta that it gets stuck and several processes remain in memory.
Just out of curiosity how much memory do you have? If you were to open ksysguard, do you see a reasonable amount of process?
Eli
On Saturday 13 March 2010 01:43:06 Neal Becker wrote:
Neal Becker wrote:
I keep getting X errors: Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
I'm starting to suspect that google chrome is the culprit.
sudo lsof -c X | grep socket | wc
219 1971 16644Now close chrome:
[nbecker@nbecker1 jupiter]$ sudo lsof -c X | grep socket | wc
60 540 4560
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
Eli Wapniarski wrote:
I'm having no such problem with Chrome. However, I have noticed, that on occaission, Chrome being Beta that it gets stuck and several processes remain in memory.
Just out of curiosity how much memory do you have? If you were to open ksysguard, do you see a reasonable amount of process?
Eli
2g memory. I've never seen this problem in 20 years of using X.
It now appears that almost certainly the problem was caused by the lastpass 1.66.11 extension of chrome. I downgraded to 1.66.10 and now sudo lsof -c X | wc appears to be stable (before it kept growing).
On Sat, 2010-03-13 at 08:04 +0200, Eli Wapniarski wrote:
I'm having no such problem with Chrome. However, I have noticed, that on occaission, Chrome being Beta that it gets stuck and several processes remain in memory.
Just out of curiosity how much memory do you have? If you were to open ksysguard, do you see a reasonable amount of process?
[Please remember not to top-post]
I use Chrome as my day-to-day browser and mostly it works well, but if for any reason it crashes, or you shut down or log out without quitting it, it tends to leave several processes lying around, as has already been said. AFAIK these processes will stay there forever unless you explicitly kill them (pgrep -fl chrome; pkill -9 chrome).
Then you often have to clean up the mess by removing stuff from ~/.cache/google-chrome/Cache and possibly ~/.config/google-chrome. I haven't completely figured it out yet, but if you don't do this then a new session of Chrome is likely to hang on some of your tabs. The exact conditions aren't clear to me.
IOW you need to remember this is Beta software, and unlike a lot of Google stuff in this case Beta really means Beta.
poc
Patrick O'Callaghan wrote:
I use Chrome as my day-to-day browser and mostly it works well, but if for any reason it crashes, or you shut down or log out without quitting it, it tends to leave several processes lying around, as has already been said. AFAIK these processes will stay there forever unless you explicitly kill them (pgrep -fl chrome; pkill -9 chrome).
Then you often have to clean up the mess by removing stuff from ~/.cache/google-chrome/Cache and possibly ~/.config/google-chrome. I haven't completely figured it out yet, but if you don't do this then a new session of Chrome is likely to hang on some of your tabs. The exact conditions aren't clear to me.
As a matter of interest, why do you use it as your "day-to-day browser" if it what seems like fairly serious deficiencies?
On Sat, 2010-03-13 at 15:25 +0000, Timothy Murphy wrote:
Patrick O'Callaghan wrote:
I use Chrome as my day-to-day browser and mostly it works well, but if for any reason it crashes, or you shut down or log out without quitting it, it tends to leave several processes lying around, as has already been said. AFAIK these processes will stay there forever unless you explicitly kill them (pgrep -fl chrome; pkill -9 chrome).
Then you often have to clean up the mess by removing stuff from ~/.cache/google-chrome/Cache and possibly ~/.config/google-chrome. I haven't completely figured it out yet, but if you don't do this then a new session of Chrome is likely to hang on some of your tabs. The exact conditions aren't clear to me.
As a matter of interest, why do you use it as your "day-to-day browser" if it what seems like fairly serious deficiencies?
The deficiencies are outweighed by the advantages, or to put it another way, by the deficiencies of other browsers. I don't want to get into a browser war here, but I though I've used FF for many years and still keep it up to date, I find Chrome extremely fast (both to start up and in page rendering) and on the whole more reliable than FF, in large part because of the process-per-tab model, which I believe FF will adopt in a future version. It's also less of a cpu hog.
The problems I mentioned above are annoying, but I know about them and they only affect me at well-defined moments.
Also, none of the above precludes me from going back to FF if/when it catches up with Chrome in these areas. Competition is good :-)
poc
Patrick O'Callaghan wrote:
On Sat, 2010-03-13 at 15:25 +0000, Timothy Murphy wrote:
Patrick O'Callaghan wrote:
I use Chrome as my day-to-day browser and mostly it works well, but if for any reason it crashes, or you shut down or log out without quitting it, it tends to leave several processes lying around, as has already been said. AFAIK these processes will stay there forever unless you explicitly kill them (pgrep -fl chrome; pkill -9 chrome).
Then you often have to clean up the mess by removing stuff from ~/.cache/google-chrome/Cache and possibly ~/.config/google-chrome. I haven't completely figured it out yet, but if you don't do this then a new session of Chrome is likely to hang on some of your tabs. The exact conditions aren't clear to me.
As a matter of interest, why do you use it as your "day-to-day browser" if it what seems like fairly serious deficiencies?
The deficiencies are outweighed by the advantages, or to put it another way, by the deficiencies of other browsers. I don't want to get into a browser war here, but I though I've used FF for many years and still keep it up to date, I find Chrome extremely fast (both to start up and in page rendering) and on the whole more reliable than FF, in large part because of the process-per-tab model, which I believe FF will adopt in a future version. It's also less of a cpu hog.
The problems I mentioned above are annoying, but I know about them and they only affect me at well-defined moments.
Also, none of the above precludes me from going back to FF if/when it catches up with Chrome in these areas. Competition is good :-)
poc
I've been using it over konq for a couple months now, cause it's blazingly fast and renders more stuff better than konq. Only within the last week I've seen this problem (a lot!). Something has suddenly changed. Maybe a recent chrome update did it.
On Friday 12 March 2010 16:43:54 Neal Becker wrote:
I keep getting X errors:
Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
I doubt if you can. This generally occurs when the site you are addressing is using a database and there are literally too many queries happening at the same moment. Sometimes it's a denial-of-service attack on the server. Sometimes it's just coincidence, and a few moments later you can connect. I have seen it regardless of web browser being used - not surprising, really.
Anne
On 03/13/2010 10:11 AM, Anne Wilson wrote:
On Friday 12 March 2010 16:43:54 Neal Becker wrote:
I keep getting X errors:
Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
I doubt if you can. This generally occurs when the site you are addressing is using a database and there are literally too many queries happening at the same moment. Sometimes it's a denial-of-service attack on the server. Sometimes it's just coincidence, and a few moments later you can connect. I have seen it regardless of web browser being used - not surprising, really.
Isn't Neal taking about X errors, not website errors?
Anne
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
On 03/13/2010 10:26 AM, Patrick Boutilier wrote:
On 03/13/2010 10:11 AM, Anne Wilson wrote:
On Friday 12 March 2010 16:43:54 Neal Becker wrote:
I keep getting X errors:
Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
I doubt if you can. This generally occurs when the site you are addressing is using a database and there are literally too many queries happening at the same moment. Sometimes it's a denial-of-service attack on the server. Sometimes it's just coincidence, and a few moments later you can connect. I have seen it regardless of web browser being used - not surprising, really.
Isn't Neal taking about X errors, not website errors?
What does xlsclients show? Got that from:
http://www.karakas-online.biz/forum/viewtopic.php?t=10494
Anne
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
On 03/13/2010 10:30 AM, Patrick Boutilier wrote:
On 03/13/2010 10:26 AM, Patrick Boutilier wrote:
On 03/13/2010 10:11 AM, Anne Wilson wrote:
On Friday 12 March 2010 16:43:54 Neal Becker wrote:
I keep getting X errors:
Maximum number of clients reachedI've never seen these before, but recently it's happening all the time. How can I debug this?
I doubt if you can. This generally occurs when the site you are addressing is using a database and there are literally too many queries happening at the same moment. Sometimes it's a denial-of-service attack on the server. Sometimes it's just coincidence, and a few moments later you can connect. I have seen it regardless of web browser being used - not surprising, really.
Isn't Neal taking about X errors, not website errors?
What does xlsclients show? Got that from:
With Intel drivers on my laptop I am getting close to the 255 max. Majority of connections being plasma-desktop:
[root@hplaptop ~]# xlsclients |wc -l && xlsclients|grep plasma-desktop|wc -l 248 178
Anne
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
kde mailing list kde@lists.fedoraproject.org https://admin.fedoraproject.org/mailman/listinfo/kde New to KDE4? - get help from http://userbase.kde.org
With Intel drivers on my laptop I am getting close to the 255 max. Majority of connections being plasma-desktop:
[root at hplaptop https://admin.fedoraproject.org/mailman/listinfo/kde ~]# xlsclients |wc -l && xlsclients|grep plasma-desktop|wc -l 248 178
You're not alone: http://bugs.kde.org/show_bug.cgi?id=230425
It seems it started with a recent Fedora update...
On Saturday 13 March 2010 14:26:17 Patrick Boutilier wrote:
On 03/13/2010 10:11 AM, Anne Wilson wrote:
On Friday 12 March 2010 16:43:54 Neal Becker wrote:
I keep getting X errors: Maximum number of clients reached
I've never seen these before, but recently it's happening all the time. How can I debug this?
I doubt if you can. This generally occurs when the site you are addressing is using a database and there are literally too many queries happening at the same moment. Sometimes it's a denial-of-service attack on the server. Sometimes it's just coincidence, and a few moments later you can connect. I have seen it regardless of web browser being used - not surprising, really.
Isn't Neal taking about X errors, not website errors?
Apologies - I should read more carefully. Googling for the error message gives guidance.
Anne
-
Anne Wilson -
Diego -
Eli Wapniarski -
Neal Becker -
Patrick Boutilier -
Patrick O'Callaghan -
Timothy Murphy