Talk-Polywell.org

Stubby wrote:i am finding it locks up less

It just locked up on me twice in a row.

I keep wondering if that 400+ page LENR thread has something to do with it. Is it possible to split the thread into two parts and lock the older part so it can't be added to anymore?

Web pages stall for me as well.

I have not been able to log on over the past week. The webpage kept giving me a 500 or 404 error.
I would get the redirect page and then the error. Mostly a 500 error.

T-P page loading continues to hang about 1/3 of the time.

When I type "www.talk-polywell.org" into the address bar it always goes immediately to a "Welcome" page which has an auto-redirect to the usual index page. But the redirect hangs with a similar frequency.

Concur with most of the comments here; also, it's worth noting that the automated error page configured in httpd.conf (usually- Apache config is so flexible that you can technically name the file most anything) is missing or inaccessible, so you're getting a 404 error as soon as the first 500 error happens. This confuses the symptoms, and may even be covering some of them up.

This is presenting like a hardware problem, most likely a wonky disk drive, or like you're out of memory and periodically the server dies and restarts after failing to get memory on a malloc/calloc/realloc() call and getting a null return.

The forum is returning several times with an error message today before the requested operation finishes.
Mostly I get this error message:

Internal Server Error

The server encountered an internal error or misconfiguration and was
unable to complete your request.

Please contact the server administrator, webmaster@talk-polywell.org and
inform them of the time the error occurred, and anything you might have
done that may have caused the error.

More information about this error may be available in the server error
log.

Additionally, a 404 Not Found error was encountered while trying to use
an ErrorDocument to handle the request.

Is it know what causes these frequent error messages?

My offer to look over the error logs and site setup stands.

A site I admin using phpbb has occasional slowdowns when the server is under load, but I haven't seen any page load failures yet.

Who exactly is the admin that would be looking into this? Today things were actually loading the first time, but taking a good while.

Joe, I have new debug info attached.

Thanks, I've noticed some seriously poor performance in recent weeks as well. On the other hand, this week it seems to be fine.

I contacted our host (Dreamhost) last week about the most recent slowdown, but they were unable to find anything wrong.

One of these days I should probably update to the latest version of php, which might bring some performance enhancements. In the meantime, I'll try to keep an eye on it. And if you have trouble, remember to send email (to admin at talk-polywell.org).

JoeStrout wrote:Thanks, I've noticed some seriously poor performance in recent weeks as well. On the other hand, this week it seems to be fine.

I contacted our host (Dreamhost) last week about the most recent slowdown, but they were unable to find anything wrong.

One of these days I should probably update to the latest version of php, which might bring some performance enhancements. In the meantime, I'll try to keep an eye on it. And if you have trouble, remember to send email (to admin at talk-polywell.org).

Are you on a virtual machine?

Do you have system administrator, probably called "root" or "sysadmin" privileges?

Do you know what operating system you're running on, at least linux or windows?

If you're root on a linux box, search for the string "oom" in your system error log. The advice is much more complex on a windows system.

Find the timestamps of the "500" responses in your error logs, then track the cause down in the DB or Apache module or system logs. That's SOP.

Trust me. I did bug fixing for Fortune 500 for a decade on Apache running on everything from VMS to AT&T UNIX to Solaris starting with version 1.2 all the way up to Solaris 10 I think is the latest one, to IBM AIX (and yes I hate SMIT) to five flavors of Linux. Oh, and windows though I don't often admit to it.

It may be something as complicated as a module told it has more memory than your virtual machine has, to something as simple as a hardset tuning parameter that is set way too low for your environment. The first thing is to get an error message that says something other than "something is broken! Ohe noes! Panic and woe!!" which is all you've seen so far.

No, this board is on a shared machine (not even a virtual private server), and of course I don't have root privs.

That obviously limits what I can do, and even what I can discover about it, somewhat. However there are a lot of things I probably can do, that I haven't tried yet. This is because I've got a lot of other demands on my time, and unless performance gets a lot worse, it's not high on my list of priorities.

Indeed, if this board turns into nothing more than a political fight club, I may as well shut it down... but I won't, because I still have reason to hope we'll eventually get something new to discuss with regard to polywell fusion!

Cheers,
- Joe

No, this board is on a shared machine (not even a virtual private server)

That explains a lot. My first experience working with "cloud" was supposed to be a virtual private machine, but experienced slowdowns from server load by other customers. The host not providing enough hardware to cover load under spikes is bad form, but I've seen it before.

By the way, who is the host? Both Amazon and Rackspace are good in my experience.

OK, the database is now denying the rights of the Apache user to grab error messages.

I need to warn you now that if you change this, you will be poking a hole in your security. It's necessary to get the data you need to fix the problem, but it's also dangerous. Don't leave it configured this way when you are not directly logged in and in charge. Hackers can steal access to your site. Be careful. Turn it off when you no longer need it. Make sure you know what you did when you turned it on, to make sure you can turn it off after you're done; take extensive notes. If you didn't and it's already done make sure the person you're working with provides you with instructions to make your site secure again.

However, take heart, it's only a small hole. All it allows, is for an attacker who has already obtained access, by another means, to edit signs of their presence from the logs. So unless you're already penetrated it's no biggie. Just make sure you don't leave it that way; some security expert like me might come along and scold you for it and whip you with a wet noodle.

Let me know if I can help.

Your site (quite rightly; it's a security precaution) will not allow me to upload the error file; I'll send it to your gmail account, if you like, but it's spam so I don't want to make a mess for you. LMK where you'd like me to post it (if you even need to know what's in it; very little, to my trained eye, but the timestamp.)

Oh, and of course it also means you got the error page fixed, which is a Very Good Thing because it's more efficient than letting the site search for something that doesn't exist. That's always longer than delivering an explicit error. A proper error is more efficient than an endless search that terminates whenever it's done finding out it has no idea what happened.

I have a couple minute-accurate timestamps and log file signatures for you to look at if you don't believe the lulzers. Stored on my text tool on my sidebar.

If you want my opinion you're on a hardware machine that is running a heavy traffic site. I suggest you review your operating agreement and see if they agreed not to screw you, and if not you move to a different provider. Good luck.