Thanks to all who replied to this problem. It set my mind at ease... till 
today.

To recap I asked why my redhat 7 (duron 700, 256 mb ram) machine was using 
so much of its memory on buffers even when it was idle. Today as I have 
gotten to a habit of doing, I did a free -t, I got something just like I 
did on wed:

wed 1/31/01 noonish
5 users on Samba. No mail running

              total       used       free     shared    buffers     cached
Mem:        257660     195060      62600      70256     137416      22812
-/+ buffers/cache:      34832     222828
Swap:       265032          0     265032
Total:      522692     195060     327632

Numbers where slightly different, but basically the same. Ok I thought this 
is fine, same as its been for the last few days. About an hour ago I 
checked again, and got this:

              total       used       free     shared    buffers     cached
Mem:        257660     254936       2724      39212     188340      38624
-/+ buffers/cache:      27972     229688
Swap:       265032       6424     258608
Total:      522692     261360     261332

There were maybe 2 active users on Samba, and not doing anything except 
printing and looking at a few documents open from the  their shares. CPU 
load was <1%. So its filled up available memory with "buffered something", 
and has now started digging into swap memory.

5 mins after this, a user informs me that their login was not working, 
another couldn't print yada yada. Basically Samba was not accepting 
connections, and existing connections were not responding. My log.smb is 
full of this over and over with nothing else of interest:

[2001/02/01 12:52:04, 0] lib/util_sock.c:write_socket_data(540)
   write_socket_data: write failure. Error = Broken pipe
[2001/02/01 12:52:04, 0] lib/util_sock.c:write_socket(566)
   write_socket: Error writing 4 bytes to socket 5: ERRNO = Broken pipe
[2001/02/01 12:52:04, 0] lib/util_sock.c:send_smb(754)
   Error writing 4 bytes to client. -1. Exiting
[2001/02/01 14:53:31, 0] smbd/service.c:make_connection(214)
   maureen (10.100.100.221) couldn't find service hp2000c:

Just made a little discovery. The last thing I did before samba stopped was 
copying the remaining files from our old server  to the Linux server.

I just did  a quick test. I made a file of about 26 mb on the windows 
desktop. Then I copied it to the share on the server. Memory usage in 
"buffers" went up by nearly exactly the same size as the file I just 
copied. Then I deleted the file, from the desktop, fileshare, and shut down 
the windows machine I was doing this on. Buffer usage is the same, and its 
not going down.

 From what I have been observing over the last few days I can convince 
myself that the increased buffer usage is the same as the files I have 
copied from old system to new.

So why are the buffers not getting cleaned up? Even after I delete the 
files the buffer usage is not going down. Over 4 days time the copied files 
in buffer, are not going away.

Any ideas on what is going on, and what I can do to fix it? What flags 
something in buffered memory as being "active" or "protected"?

Thanks for any help again.
-Alex

Current ps and update -d below

===============================
update -d

bdflush version 1.4
0:    40 Max fraction of LRU list to examine for dirty blocks
1:   500 Max number of dirty blocks to write each time bdflush activated
2:    64 Num of clean buffers to be loaded onto free list by refill_freelist
3:   256 Dirty block threshold for activating bdflush in refill_freelist
4:   500 Percentage of cache to scan for free clusters
5:  3000 Time for data buffers to age before flushing
6:   500 Time for non-data (dir, bitmap, etc) buffers to age before flushing
7:  1884 Time buffer cache load average constant
8:     2 LAV ratio (used to determine threshold for buffer fratricide).

===============================
A ps -ax on the fresh boot shows the following for reference:

   PID TTY      STAT   TIME COMMAND
     1 ?        S      0:05 init [3]
     2 ?        SW     0:00 [kflushd]
     3 ?        SW     0:00 [kupdate]
     4 ?        SW     0:00 [kpiod]
     5 ?        SW     0:00 [kswapd]
     6 ?        SW<    0:00 [mdrecoveryd]
   319 ?        S      0:00 syslogd -m 0
   329 ?        S      0:00 klogd
   344 ?        S      0:00 portmap
   360 ?        SW     0:00 [lockd]
   361 ?        SW     0:00 [rpciod]
   371 ?        S      0:00 rpc.statd
   423 ?        S      0:00 identd -e -o
   426 ?        S      0:00 identd -e -o
   427 ?        S      0:00 identd -e -o
   434 ?        S      0:00 identd -e -o
   435 ?        S      0:00 identd -e -o
   442 ?        S      0:00 /usr/sbin/atd
   457 ?        S      0:00 xinetd -reuse -pidfile /var/run/xinetd.pid
   466 ?        S      0:00 /usr/sbin/sshd
   487 ?        S      0:00 lpd Waiting
   531 ?        S      0:00 sendmail: accepting connections
   547 ?        S      0:00 gpm -t ps/2
   689 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   702 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   703 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   704 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   705 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   706 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   707 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   708 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   709 ?        S      0:00 /usr/sbin/httpd -D HAVE_PERL -D HAVE_PROXY -D 
HAVE_SS
   712 ?        S      0:00 crond
   745 ?        S      0:00 xfs -droppriv -daemon
   775 ?        S      0:00 rhnsd --interval 120
   805 ?        S      0:00 smbd -D
   815 ?        S      0:00 nmbd -D
   826 tty2     S      0:00 /sbin/mingetty tty2
   827 tty3     S      0:00 /sbin/mingetty tty3
   828 tty4     S      0:00 /sbin/mingetty tty4
   829 tty5     S      0:00 /sbin/mingetty tty5
   830 tty6     S      0:00 /sbin/mingetty tty6
   833 ?        S      0:00 perl /usr/libexec/webmin/miniserv.pl 
/etc/webmin/mini
   889 ?        S      0:00 smbd -D
   890 ?        S      0:00 CROND
   891 ?        S      0:00 bash /usr/bin/run-parts /etc/cron.hourly
   901 ?        S      0:00 awk -v progname=/etc/cron.hourly/sysstat 
progname {??
   902 ?        S      0:00 sh /usr/lib/sa/sa1 600 6
   904 ?        S      0:00 /usr/lib/sa/sadc 600 6 /var/log/sa/sa01
   943 ?        S      0:00 smbd -D
   946 ?        S      0:00 smbd -D
   964 tty1     S      0:00 /sbin/mingetty tty1
   998 ?        S      0:00 smbd -D
  1052 ?        S      0:00 in.telnetd: 
troll
  1053 pts/0    S      0:00 login -- alex
  1054 pts/0    S      0:00 -bash
  1251 pts/0    R      0:00 ps -ax



-Alex Tabony - [EMAIL PROTECTED]
510/849-2911 Voice/TTY
510/849-2968 Fax

Not all those who wander are lost. -jrrt



_______________________________________________
Redhat-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/redhat-list

Reply via email to