Linux Community's Journal|
[Most Recent Entries]
Tuesday, September 24th, 2002
i'm pretty sure my machine was hacked overnight. i woke up this morning to find my mysql server no longer running, and a whole bunch of kernel "out of memory" errors in my syslog. in addition, i have several messages like this one:Sep 24 05:01:31 machinename kernel: UDP: bad checksum. From some.random.ip.address:4156 to my.ip.address.here:4156 ulen 49
any idea what would cause all of these "bad checksum" messages? i don't know what to patch if i don't know what got broken :P
thanks in advance. Current Mood: concerned
I currently have a colocated server running FreeBSD and I've been contemplating switching it to Debian, but the problem is, I don't have direct access to it (unless I go into the facility).
Crazy idea that probably isn't possible, but could you like mount a ramdisk and run an installer off that? Hmm, this is probably completely implausible, but I'd hate to go in and bother the staff at the colo facility.
Wondering if anyone knows of some good logistics software for the managament of logs.
Currently in /var/log I have an authorization log and a few specialized system logs that I have made to custom fit the system.
I am using a really awful perl script running as a cron job to tar a log after a month. I am running into issues with space, as the logs
I have are now really starting to accumulate - this server only has three gigabytes of storage. This being it would be rather handy to grep undeeded non-pertainent mass of my logs. I was thinking about inscructing the machine to ignore certain things, however, I given rule sets far to broad - I do not want to ignore that which is pertainent. This being I thought a perl script that would grep out any redundancy/non-pertainent mass after the thirty day peroid would be much more effective. In my tests I have ended up greping out things to which I would have rather kept, and all together failing. Any more effcient programers amoung us have a solution implemented for such an issue? Thanks in advance.
|Debian Kernel Question
I recently joined the enlightened and switched to Debian. I downloaded a minimal iso and opted for the network install.
I'm having trouble getting the system to recognize usual stuff like cd-writer, usb-storage etc. So I decided to upgrade the kernel and compile in what I wanted.
I downloaded kernel headers and source for 2.4.19. copied it to /usr/src. Ran `make mrproper',`make xconfig', `make dep', `make bzImage' `make modules' and `make modules_install'. Symlinked the image to /boot, updated lilo to the symlink in boot and reran lilo. I even tried copying the image itself to /boot ... nada.
After rebooting, the system hangs. It won't come all the way up. Get's to `klogd' and stops. Did I miss something? Or does Debian handle installing new kernels differently than what I'm used to?
Why is Pan such a complete pain in the ass to use?
Example 1 To ignore a thread:
In Agent Press 'I'
In Pan it's Ctrl-Shift-I
Why so much more complicated for a common event? When Pan does ignore a thread, it stays in the count, so you don't see from the summary that there is in fact nothing to see. It will also chunter away for a few seconds with the screen flashing while it actually marks them ignored.
In Agent, it just ignores them.
Agent will also delete, or mark as read the line, and any other parts of the thread that is collapsed. In Pan, you have to delete the first, then go back and delete the rest.
And don't even get me started on email readers.... Current Mood: aggravated