|
From: Micah Y. <yo...@ho...> - 2001-06-11 19:04:53
|
OK, this is Screwed-Up(tm). Finally got my co-located box for Slash (and other things). Now, the slashd process is pretty much running out of control. It's in the R state, taking 99% of a CPU (there are two). Now, whenever I try to access anything in the Slash site's log directory (containing slashd.log), the process (ls, mc, less, etc.) hangs and is in a D state. Whenever I try to strace any of those processes, the same fate happens to the strace process. I can't kill -9 any of them!!! I know you can't kill a D process but why not the R slashd? In fact, running reboot, the same thing happened to the shutdown process! Aaargh! It's on an XFS filesystem. I wonder if that has anything to do with it. Any ideas for how to get out of this mess? The box is a 6-hour drive away and they charge $25 for power cycling, so I'd only want to use that as a last resort. Thanks, Micah |
|
From: Brian A. <br...@ta...> - 2001-06-11 19:16:39
|
Micah Yoder wrote:
> Finally got my co-located box for Slash (and other things). Now, the slashd
> process is pretty much running out of control. It's in the R state, taking
> 99% of a CPU (there are two).
grep slash | awk '{print $2}' | xargs kill -9
That normally solves it for me :)
-Brian
|
|
From: Nomad M. L. <er...@no...> - 2001-06-11 19:40:48
|
OK, After much fighting, and manual reading, I got slash to install. But when I hit the site, all I get is a blank page. The source of which reads: <!-- start template: ID , header;; --> <!-- end template: ID , header;; --> <!-- start template: ID , index;; --> <!-- end template: ID , index;; --> and that's it. Nothing else. Can anyone clue me into where I should start in trying to fix this? Thanks. --Eric |
|
From: Alvaro d. C. <ac...@ba...> - 2001-06-12 09:22:32
|
El 11 Jun 2001 15:40:45 -0400, Nomad Media Lab escribió: > > OK, > > After much fighting, and manual reading, I got slash to install. But when > I hit the site, all I get is a blank page. The source of which reads: > > > <!-- start template: ID , header;; --> > > <!-- end template: ID , header;; --> > > <!-- start template: ID , index;; --> > > <!-- end template: ID , index;; --> > > and that's it. > Nothing else. > > Can anyone clue me into where I should start in trying to fix this? It seems that the templates are not in the database. Can you check it? Cheers -- Alvaro > > Thanks. > > --Eric > > > _______________________________________________ > Slashcode-general mailing list > Sla...@li... > http://lists.sourceforge.net/lists/listinfo/slashcode-general > |
|
From: Nomad M. L. <er...@no...> - 2001-06-12 17:34:35
|
> > > > OK, > > > > After much fighting, and manual reading, I got slash to install. But when > > I hit the site, all I get is a blank page. The source of which reads: > > > > > > <!-- start template: ID , header;; --> > > > > <!-- end template: ID , header;; --> > > > > <!-- start template: ID , index;; --> > > > > <!-- end template: ID , index;; --> > > > > and that's it. > > Nothing else. > > > > Can anyone clue me into where I should start in trying to fix this? > > It seems that the templates are not in the database. Can you check it? > > Cheers > > -- Alvaro > Thanks. I checked and somehow all the templates are not in the database. Some are, some are not. If I call /search.pl I get part of a page, as the storysearch;search;default template is there, but most of the others are not). What is the best way to recreate the templates database table without having to reinstall the slashsite? or is this a symptom of a larger problem and I should start over from the beginning? Thanks again for the help. --Eric |
|
From: Alvaro d. C. <ac...@ba...> - 2001-06-12 19:10:00
|
El 12 Jun 2001 13:34:32 -0400, Nomad Media Lab escribió: > > > > > > OK, > > > > > > After much fighting, and manual reading, I got slash to install. But when > > > I hit the site, all I get is a blank page. The source of which reads: > > > > > > > > > <!-- start template: ID , header;; --> > > > > > > <!-- end template: ID , header;; --> > > > > > > <!-- start template: ID , index;; --> > > > > > > <!-- end template: ID , index;; --> > > > > > > and that's it. > > > Nothing else. > > > > > > Can anyone clue me into where I should start in trying to fix this? > > > > It seems that the templates are not in the database. Can you check it? > > > > Cheers > > > > -- Alvaro > > > > Thanks. > I checked and somehow all the templates are not in the database. Some are, > some are not. If I call /search.pl I get part of a page, as the > storysearch;search;default template is there, but most of the others are > not). > What is the best way to recreate the templates database table without > having to reinstall the slashsite? or is this a symptom of a larger > problem and I should start over from the beginning? Yes, this is a bad start. Maybe I try to install all from scratch. It is very quickly. You can use the template-tool tool to insert very easly the templates in the database. Good Luck! -- Alvaro > > Thanks again for the help. > > > --Eric > > > _______________________________________________ > Slashcode-general mailing list > Sla...@li... > http://lists.sourceforge.net/lists/listinfo/slashcode-general > |
|
From: Nomad M. L. <er...@no...> - 2001-06-12 19:27:30
|
Thanks for the help. I now have two slash sites running on one machine, now comes the fun part. --Eric On Mon, 11 Jun 2001, Nomad Media Lab wrote: > Date: Mon, 11 Jun 2001 15:40:45 -0400 (EDT) > From: Nomad Media Lab <er...@no...> > To: sla...@li... > Subject: newbie question > > > OK, > > After much fighting, and manual reading, I got slash to install. But when > I hit the site, all I get is a blank page. The source of which reads: > > > <!-- start template: ID , header;; --> > > <!-- end template: ID , header;; --> > > <!-- start template: ID , index;; --> > > <!-- end template: ID , index;; --> > > and that's it. > Nothing else. > > Can anyone clue me into where I should start in trying to fix this? > > Thanks. > > --Eric > > --------------------------------- Nomad Media Lab design + technology consulting http://www.nomadlab.com |
|
From: Micah Y. <yo...@ho...> - 2001-06-11 19:29:15
|
> grep slash | awk '{print $2}' | xargs kill -9
>
> That normally solves it for me :)
> -Brian
uhh... first of all what are you grepping?
Second, I *did* kill -9 the slashd process, several times. And it's still
running!
And why is every process that accesses the log directory turning into a D?
Thanks...
|
|
From: Brian A. <br...@ta...> - 2001-06-11 20:01:44
|
Micah Yoder wrote:
>
> > grep slash | awk '{print $2}' | xargs kill -9
> >
> > That normally solves it for me :)
> > -Brian
>
> uhh... first of all what are you grepping?
slash (aka the daemon that is running).
> Second, I *did* kill -9 the slashd process, several times. And it's still
> running!
Right now the above awk piece works for me.
> And why is every process that accesses the log directory turning into a D?
Not sure to be honest. This is happening with me on RedHat 7.1, but not
6.2.
-Brian
|
|
From: Micah Y. <yo...@ho...> - 2001-06-11 20:17:22
|
> > uhh... first of all what are you grepping? > > slash (aka the daemon that is running). hmm. 'grep slash' reads stdin. Did you mean to pipe ps into it? >> And why is every process that accesses the log directory turning into a D? > Not sure to be honest. This is happening with me on RedHat 7.1, but not > 6.2. Crap. So I'm not the only one it's hitting. That is really strange. What in the world could slashd be doing to make every process that hits its log directory hang? (No, I haven't yet extensively studied the code...) My test system at home has never done that. But I don't leave it on all night. This started happening sometime last night on my server, after it was up a day and a half. You're not running on XFS are you? |
|
From: Brian A. <br...@ta...> - 2001-06-11 20:23:54
|
Micah Yoder wrote: > hmm. 'grep slash' reads stdin. Did you mean to pipe ps into it? Woops, add a ps auwx | before the beginning of that. > >> And why is every process that accesses the log directory turning into a D? > > Not sure to be honest. This is happening with me on RedHat 7.1, but not > > 6.2. > Crap. So I'm not the only one it's hitting. That is really strange. What > in the world could slashd be doing to make every process that hits its log > directory hang? (No, I haven't yet extensively studied the code...) Not sure, I can kill mine off with ps so I haven't made a big deal of ti yet (and this is not happening under Debian, whish is what slashcode and use.perl use for production machines). > You're not running on XFS are you? Nope, wish I was though :) -Brian |
|
From: Micah Y. <yo...@ho...> - 2001-06-12 00:02:32
|
> Not sure, I can kill mine off with ps so I haven't made a big deal of > ti yet (and this is not happening under Debian, whish is what slashcode > and use.perl use for production machines). hmm. Maybe it's a different problem if you can kill yours. Mine was completely unkillable and made all processes accessing that log directory hang. > > You're not running on XFS are you? > > Nope, wish I was though :) If my problem was an XFS bug then thank your lucky stars you're not running it! |
|
From: Brian A. <br...@ta...> - 2001-06-12 00:18:43
|
Micah Yoder wrote: > hmm. Maybe it's a different problem if you can kill yours. Mine was > completely unkillable and made all processes accessing that log directory > hang. What kernel are you on? I seem to remember something about a bug like this in the 2.4.2 kernels (I have upgraded to 2.4.3). -Brian |
|
From: Micah Y. <yo...@ho...> - 2001-06-12 00:26:33
|
On Monday 11 June 2001 20:12, you wrote: > Micah Yoder wrote: > > hmm. Maybe it's a different problem if you can kill yours. Mine was > > completely unkillable and made all processes accessing that log directory > > hang. > > What kernel are you on? I seem to remember something about a > bug like this in the 2.4.2 kernels (I have upgraded to 2.4.3). > -Brian Running the kernel that SGI supplied on their CD. Basically it's the Red Hat 7.1 kernel, which is 2.4.2 plus a whole bunch of patches that Red Hat claims "fixes things". Now I'm a bit nervous to run slashd again. :-) I guess I should re-compile the kernel with 2.4.5 + CVS XFS. That could be tricky though, and I'm worried about virtual memory problems in later kernels. |