fedora 1 INCREDIBLY slow
 

over the last few weeks it's gotten much slower, like 10-15 seconds to open an xterm window slow. i think it's because i have a ton of processes taking up memory somehow, free says

and i ran ps-e and it says i've got all this stuff going too

all i've got open is firefox, firebird, gaim, and xmms, as well as the xterm i ran that in. seems like there's a log going on, and a lot of repeat processes on the list. any help?

my opinion:

too many daemons brought up during startup. try executing the command "ntsysv" and uncheck those processes that aren't needed.

As root run the top command. Then hit SHIFT-M. That will sort processes by memory usage and show you processes with the highest memory usage first. SHIFT-P will show you process with the most processor usage first.
I noticed a couple xmms processes along with firefox and thunderbird. I've had mozilla go nuts on me in the past and start using 98% of my processor and up to 80% of my memory. If something like xmms, firefox, or thunderbird are chewing up all of your memory or processor kill off the process. You can easily do that by pressing k while top is running. It will then prompt you for a process id to kill. Look in the PID column for the correct number to enter. If you decide not to kill a process just press enter and the PID to kill: prompt will go away and take you back to top. Press q to quit the top program.

"ntsysv" is not a command i have??

none of the processes are using a great ammount of memory by themselves, just X, firefox and thunderbird are big, the rest are under 2%, but all the kdeinits are almost 2%each... so it's not like there's one program doing it all

When you run top what do the load average numbers show in the upper right corner?

I was seeing a similar performance problem that was only affecting some X programs. The symptom was that they would start VERY slowly. And during bootup, sendmail was starting VERY slowly as well.

Try adding ".localdomain" versions of each host name in your /etc/hosts file.
BEFORE
127.0.0.1 localhost
192.168.1.1 firewall
192.168.1.102 bobby

AFTER
127.0.0.1 localhost.localdomain localhost
192.168.1.1 firewall.localdomain firewall
192.168.1.102 bobby.localdomain bobby

This worked for me. Sendmail starts very quickly now (on RedHat 8 and 9 systems).

As a replacement to tksysv you now have chkconfig and service

in /etc/rc.d/init.d you have scripts that can be setup to start in certain runlevels. You can control them with chkconfig.

chkconfig smb off
service smb stop

The service command starts and stops services, chkconfig sets up the scripts to start and stop on boot or runlevel changes.

Also chkconfig controls the daemons in /etc/xinetd.d, they are not services so they cannot be controlled by service. chkconfig will restart xinetd to stop or start them as needed.

However it does sound like it's just your /etc/host file causing slow startups. The ram usage can be misleading since you are looking at cached and buffered ram as well as ram that's actually used. The processes you show running are not looking very abnormal. I think you'll be ok on your memory as is.

that's exactly what's happening to me, but when i went to change my etc/hosts there's only one line it it, and it already has the localdoman in front.. so i don't know what to do with it.

thanks for all the help so far!

Well, several things have to be in sync. Here's the portions of my system configuration that I think are important in this respect. There are three files: /etc/hosts, /etc/sysconfig/network, and /etc/sysconfig/network-scripts/ifcfg-eth0

My /etc/hosts contains:
# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1 localhost.localdomain localhost
192.168.1.1 firewall.localdomain firewall
192.168.1.10 gort.localdomain gort
192.168.1.11 klaatu.localdomain klaatu
192.168.1.102 bobby.localdomain bobby
192.168.1.201 printer.localdomain printer

My /etc/sysconfig/network contains:
NETWORKING=yes
HOSTNAME=bobby

And my /etc/sysconfig/network-scripts/ifcfg-eth0 contains:
DEVICE=eth0
ONBOOT=yes
BOOTPROTO=dhcp

I was told that the /etc/sysconfig/network file *should not* specify the domain name on the same line as the hostname. When this is done, the system has to figure out the domain from somewhere else. If it can't find it, then it assumes ".localdomain" but only uses the fully-qualified name "bobby.localdomain" (in my case) a couple of times. During sendmail's startup and during the launch of some (but not all) X programs. When the ".localdomain" entries are missing in /etc/hosts, the system tries to resolve the full name through DNS and that is what takes several minutes (to fail).

Good Luck!

here's all 3 of them, they seem to be similar to yours now, but i only have 1 line in hosts.


/etc/hosts
/etc/sysconfig/network
/etc/sysconfig/network-scripts/ifcfg-eth0

Okay, I'm pretty sure you need to add a line to your /etc/hosts file that has the IP address your system is being assigned, and your hostname with ".localdomain" and then the hostname again. The only thing you need to find out is your IP address, and that's easy to get.
In a shell (terminal), type the following command and press enter:

/sbin/ifconfig

Here's what I get when I do it:

[edski@bobby edski]$ /sbin/ifconfig
eth0 Link encap:Ethernet HWaddr 00:0D:56:EB:3D:B5
inet addr:192.168.1.102 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1356 Metric:1
RX packets:143582 errors:0 dropped:0 overruns:0 frame:0
TX packets:313549 errors:0 dropped:0 overruns:0 carrier:0
collisions:91975 txqueuelen:1000
RX bytes:14227779 (13.5 Mb) TX bytes:278699272 (265.7 Mb)
Interrupt:11

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:70457 errors:0 dropped:0 overruns:0 frame:0
TX packets:70457 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:4803013 (4.5 Mb) TX bytes:4803013 (4.5 Mb)

The section about "eth0" is the area of interest, and the "inet addr" is the value that's needed. In my case, my IP address is 192.168.1.102 and, if that's the address your system is assigned, here's what /etc/hosts should look like:

127.0.0.1 localhost.localdomain localhost
192.168.1.102 theone.localdomain theone

If your address is something other than 192.168.1.102, then put the appropriate address in that place.

I think you'll need to reboot your system after making the change -- and I think you'll see a dramatic difference.

Good Luck!!!

wow, that worked great! and i didn't have to restart! thanks a lot for all your help

Glad to help. It took me a loooong time to find the answer (and I finally found it here at LinuxQuestions.org) so I figured I'd pay back the favor and post my solution to several queries that looked like they might be having the same problem. As it turns out, that's exactly what happened in your case. I'm pleased it worked out so well.

Feel free to send me an Email message (there's a button around here somewhere that'll let you do that).

See ya!

I'm having similar problems, except I can't seem to fix it.
I remember this happening a while ago, and I fixed it then, but I don't remember what I did.

This time around, problems cropped up when I changed the hostname on my linux box (FC3 BTW)
However, I can't seem to make the problem go away (xterm takes about 2 minutes to load)

Any thoughts? Thanks in advance!

Here are my files:

#/etc/hosts
# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1 localhost.localdomain localhost frodo
192.168.0.101 localhost.localdomain frodo.localdomain frodo localhost
192.168.0.102 bilbo.localdomain bilbo
192.168.0.103 sam.localdomain sam

#/etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE=eth0
ONBOOT=yes
BOOTPROTO=none
IPADDR=192.168.0.101
NETMASK=255.255.255.0
GATEWAY=192.168.0.1
TYPE=Ethernet

#/etc/sysconfig/network
NETWORKING=yes
HOSTNAME=frodo

Drop at least the red coloured parts.