Re: [cobalt-users] RaQ3 dies again for no reason - we maybe no reason

[craig <craig@xxxxxxxxxxxxx>]

I had a couple of servers doing this to me are there cron running at the
time it happens like running reports and analog ?
This can cuase that you can disbale analog if you do not need it

If your load avaerges are going through the roof you will not get a prompt
when you try to ssh does it try to  connect eg just appears to hang and
maybe does not give a prompt ?

That indicates high load also grep .cgi /var/log/httpd
maybe a users cgi went out of control I have had that I made a new cgiwrap
with a max excution time and max memory

If asp running lower the script timouts

I also had another box with faulty drive frist thought it was the ide
cable but changed out the drive and cable  that fixed that

maybe a fauly ethernet card off course that means changing out the box as
its built in you can get an inication of that if the collisons are high

problay means faulty card or it did not start in 100mhz mode

These are just some ideas





On Tue, 6 Aug 2002, Chae wrote:

> Hi Yah,
>
> Me again :)  The RaQ3 that died for no reason last week died again tonight.
>
> On the logs an hour before it died the following could be seen...
> Aug 6 19:18:34 ns kernel: possible SYN flooding on port 80. Sending cookies.
>
> Then an hour later the same as last week - no access via HTTP, FTP or SSH
> yet the logs prior to reboot were as follows:  xxx.xxx being our IP