PDA

View Full Version : load v cpu usage


chrispadfield
Sat 2nd Jun '01, 5:21pm
I am very confused by this. The load is meant to give an indicationn of how the server is running with a high number bad.

Thing is, how do i get loads of above 10 with 60% of the cpu not being used and spare ram as well. I don't understand that.

chrispadfield
Sun 3rd Jun '01, 10:10am
does no one know the answer to this? it is unusal for this to happen or is there something wrong with the question?

again my server load is 7.8 but idle% has not fallen below 70. What is going on!

chrispadfield
Sun 3rd Jun '01, 10:20am
ok i found something in usenet and it suggested something else will be holding up the machine, something that takes time but does not use the processor. They suggested disk writing might be the case.

So...

i did ps aux to see what was there. sorry for the big list:


USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
root 1 0.0 0.0 1100 48 ? S Jun02 0:05 init [3]
root 2 0.0 0.0 0 0 ? SW Jun02 0:01 [kflushd]
root 3 0.0 0.0 0 0 ? DW Jun02 0:00 [kupdate]
root 4 0.0 0.0 0 0 ? SW Jun02 0:00 [kpiod]
root 5 0.0 0.0 0 0 ? SW Jun02 0:24 [kswapd]
root 6 0.0 0.0 0 0 ? SW< Jun02 0:00 [mdrecoveryd]
root 277 0.0 0.0 1096 0 ? SW Jun02 0:00 [apmd]
root 330 0.0 0.0 1156 152 ? S Jun02 0:01 syslogd -m 0
root 341 0.0 0.0 1416 160 ? S Jun02 0:00 klogd
daemon 357 0.0 0.0 1128 0 ? SW Jun02 0:00 [atd]
root 373 0.0 0.0 1304 28 ? S Jun02 0:00 crond
root 389 0.0 0.0 1120 0 ? SW Jun02 0:00 [inetd]
named 405 0.0 0.3 3560 900 ? S Jun02 0:18 named -u named
root 421 0.0 0.0 1176 0 ? SW Jun02 0:00 [lpd]
qmails 431 0.0 0.0 1132 124 ? S Jun02 0:05 qmail-send
qmaill 432 0.0 0.0 1084 20 ? S Jun02 0:00 /usr/local/bin/cyclog /var/log/qmail
root 433 0.0 0.0 1136 0 ? SW Jun02 0:00 [tcpserver]
qmails 434 0.0 0.0 1136 12 ? S Jun02 0:00 tcpserver -H -R -x /etc/tcp.smtp.cdb -c100 -u532 -g527 0 s
qmaill 441 0.0 0.0 1068 24 ? S Jun02 0:01 /usr/local/bin/accustamp
root 442 0.0 0.0 1088 72 ? S Jun02 0:00 qmail-lspawn ./Maildir/
qmailr 443 0.0 0.0 1088 0 ? SW Jun02 0:00 [qmail-rspawn]
qmailq 444 0.0 0.0 1080 40 ? S Jun02 0:00 qmail-clean
root 598 0.0 0.0 4936 148 ? S Jun02 0:01 httpd
root 606 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 607 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 608 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 609 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 610 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 611 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 612 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 614 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 615 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 616 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 617 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 618 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 619 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 620 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 621 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 623 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 624 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 625 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 626 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 627 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 634 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 635 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 636 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 637 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 638 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 639 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 640 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 641 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 642 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 643 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 644 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 645 0.0 0.0 1664 4 ? S Jun02 0:00 [rotatelogs]
root 646 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 651 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 652 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 653 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 654 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 655 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 656 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 657 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 658 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 659 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 660 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 661 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 662 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 663 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 664 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 665 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 666 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 667 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 668 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 669 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 670 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 671 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 672 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 673 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 674 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 675 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 676 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 678 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 679 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 680 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 681 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 682 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 683 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 684 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 685 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 686 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 687 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 688 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 689 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 690 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 691 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 692 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 693 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 694 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 695 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 696 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 698 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 699 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 700 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 701 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 702 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 703 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 704 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 705 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 706 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 707 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 708 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 709 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 710 0.0 0.0 1664 20 ? S Jun02 0:00 [rotatelogs]
root 711 0.0 0.0 1664 20 ? S Jun02 0:01 [rotatelogs]
root 712 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 713 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 714 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 715 0.0 0.0 1664 0 ? SW Jun02 0:00

chrispadfield
Sun 3rd Jun '01, 10:20am
[rotatelogs]
root 716 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 717 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 718 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 726 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 727 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 728 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 729 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 730 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 731 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 732 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 733 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 734 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 735 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 736 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 737 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 738 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 739 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
root 750 0.0 0.0 4380 44 ? S Jun02 0:01 httpsd -DSSL
nobody 753 0.0 0.0 2448 0 ? SW Jun02 0:00 [gcache]
root 769 0.0 0.0 1076 0 tty1 SW Jun02 0:00 [mingetty]
root 770 0.0 0.0 1076 0 tty2 SW Jun02 0:00 [mingetty]
root 771 0.0 0.0 1128 0 ttyS0 SW Jun02 0:00 [getty]
nobody 774 0.0 0.0 4556 0 ? SW Jun02 0:00 [httpsd]
nobody 775 0.0 0.0 4556 0 ? SW Jun02 0:00 [httpsd]
nobody 776 0.0 0.0 4556 0 ? SW Jun02 0:00 [httpsd]
nobody 777 0.0 0.0 4556 0 ? SW Jun02 0:00 [httpsd]
nobody 778 0.0 0.0 4556 0 ? SW Jun02 0:00 [httpsd]
admin 874 0.0 0.0 1756 0 ? SW Jun02 0:00 [bash]
root 885 0.0 0.0 2052 0 ? SW Jun02 0:00 [su]
root 887 0.0 0.0 1756 0 ? SW Jun02 0:00 [bash]
root 899 0.0 0.0 1688 0 ? SW Jun02 0:00 [safe_mysqld]
root 910 0.0 5.6 63144 14492 ? SN Jun02 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 915 0.0 5.6 63144 14492 ? SN Jun02 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 916 0.0 5.6 63144 14492 ? SN Jun02 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 10569 0.0 0.0 1664 0 ? SW Jun02 0:00 [rotatelogs]
nobody 24917 0.3 1.9 14060 4964 ? D 03:44 0:15 httpd
root 24939 0.1 5.6 63144 14492 ? SN 03:44 0:07 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 24961 0.3 1.2 14056 3124 ? S 03:45 0:14 httpd
nobody 24962 0.4 3.8 14200 9944 ? S 03:45 0:17 httpd
root 24984 0.1 5.6 63144 14492 ? SN 03:46 0:04 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 24986 0.4 1.9 14040 5044 ? D 03:46 0:17 httpd
root 24987 0.2 5.6 63144 14492 ? DN 03:46 0:11 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 25044 0.2 1.6 14020 4288 ? S 03:47 0:08 httpd
nobody 25045 0.2 3.4 14156 8872 ? S 03:47 0:08 httpd
root 25066 0.2 5.6 63144 14492 ? SN 03:48 0:08 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 25127 0.1 5.6 63144 14492 ? SN 03:49 0:06 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 25210 0.0 5.6 63144 14492 ? SN 03:52 0:02 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 25221 0.3 1.9 14016 4972 ? S 03:53 0:13 httpd
nobody 25322 0.3 2.6 14032 6892 ? S 03:54 0:12 httpd
root 25393 0.0 5.6 63144 14492 ? SN 03:56 0:02 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 25574 0.0 5.6 63144 14492 ? SN 04:00 0:02 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 25772 0.3 2.0 14164 5204 ? S 04:04 0:10 httpd
root 25873 0.1 5.6 63144 14492 ? SN 04:05 0:03 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 26070 0.4 1.3 14012 3352 ? S 04:10 0:11 httpd
root 26102 0.2 5.6 63144 14492 ? DN 04:11 0:05 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 26196 0.3 1.8 14012 4660 ? S 04:14 0:08 httpd
nobody 26197 0.2 0.9 7560 2484 ? S 04:14 0:06 httpd
root 26271 0.1 5.6 63144 14492 ? SN 04:15 0:02 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 26311 0.0 5.6 63144 14492 ? SN 04:16 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 26877 0.4 2.7 13776 7196 ? S 04:28 0:08 httpd
root 26934 0.2 5.6 63144 14492 ? SN 04:29 0:04 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27038 0.1 3.2 13780 8268 ? S 04:31 0:02 httpd
nobody 27039 0.6 2.7 14028 6996 ? S 04:31 0:08 [httpd]
root 27067 0.4 5.6 63144 14492 ? DN 04:33 0:06 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27105 0.1 5.6 63144 14492 ? SN 04:33 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27317 0.2 1.0 7528 2772 ? S 04:38 0:02 httpd
nobody 27322 0.3 2.3 13952 6100 ? S 04:38 0:03 httpd
nobody 27352 0.3 0.1 13976 508 ? S 04:38 0:03 httpd
root 27356 0.2 5.6 63144 14492 ? SN 04:39 0:02 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27403 0.5 5.6 63144 14492 ? SN 04:39 0:05 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27458 0.0 5.6 63144 14492 ? DN 04:41 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27506 0.2 1.4 8396 3708 ? S 04:42 0:02 httpd
nobody 27507 0.1 1.0 7352 2596 ? S 04:42 0:01 httpd
root 27533 0.2 5.6 63144 14492 ? SN 04:43 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27570 0.2 5.6 63144 14492 ? DN 04:44 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27602 0.2 1.2 13688 3180 ? S 04:44 0:01 httpd
root 27629 0.0 5.6 63144 14492 ? SN 04:44 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27663 0.0 0.0 1704 64 ? S 04:45 0:00 in.telnetd
root 27684 0.0 0.0 2280 0 pts/1 SW 04:46 0:00 [login]
admin 27691 0.0 0.0 1756 0 pts/1 SW 04:46 0:00 [bash]
root 27703 0.0 0.0 2052 0 pts/1 SW 04:46 0:00 [su]
root 27712 0.0 0.1 1756 424 pts/1 S 04:46 0:00 bash
nobody 27734 0.1 3.9 13764 10304 ? D 04:46 0:01 httpd
nobody 27739 0.2 3.4 13772 9020 ? S 04:47 0:01 httpd
nobody 27742 0.0 0.9 7140 2392 ? S 04:47 0:00 httpd
nobody 27743 0.1 0.9 7388 2560 ? S 04:47 0:00 httpd
nobody 27744 0.0 0.9 7424 2536 ? S 04:47 0:00 httpd
nobody 27745 0.1 1.2 7144 3312 ? S 04:47 0:00 httpd
nobody 27746 0.2 3.4 13648 8904 ? S 04:47 0:01 httpd
nobody 27747 0.2 1.1 7368 3088 ? S 04:47 0:01 httpd
root 27750 0.6 5.6 63144 14496 ? SN 04:47 0:03 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27751 0.7 5.6 63144 14496 ? SN 04:47 0:03 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27753 0.1 5.6 63144 14496 ? DN 04:47 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27755 0.7 5.6 63144 14496 ? DN 04:47 0:03 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27757 0.2 3.4 13852 8876 ? S 04:47 0:01 httpd
root 27758 0.7 5.6 63144 14496 ? SN 04:47 0:03 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27791 0.5 5.6 63144 14496 ? SN 04:48 0:02 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27802 0.1 5.6 63144 14496 ? SN 04:48 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27830 0.1 5.6 63144 14496 ? SN 04:49 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27854 0.4 3.2 13804 8280 ? S 04:50 0:01 httpd
nobody 27857 0.3 3.4 13732 8880 ? S 04:50 0:00 httpd
nobody 27861 0.2 1.1 7388 2976 ? S 04:50 0:00 httpd
root 27878 0.0 5.6 63144 14496 ? SN 04:51 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27893 0.0 5.6 63144 14496 ? SN 04:52 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27896 0.1 5.6 63144 14496 ? DN 04:52 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27907 0.3 1.1 7052 2880 ? S 04:53 0:00 httpd
nobody 27911 0.2 1.1 7416 2956 ? S 04:53 0:00 httpd
nobody 27913 0.7 4.1 13728 10632 ? S 04:53 0:01 httpd
nobody 27916 0.1 0.8 7216 2296 ? S 04:53 0:00 httpd
nobody 27917 0.1 0.8 7200 2272 ? S 04:53 0:00 httpd
nobody 27918 0.4 3.3 13600 8756 ? S 04:53 0:00 httpd
nobody 27921 0.0 0.0 5024 100 ? S 04:53 0:00 httpd
root 27923 0.1 5.6 63144 14496 ? SN 04:53 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27928 0.0 5.6 63144 14496 ? SN 04:53 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27931 0.0 5.6 63144 14496 ? SN 04:53 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27934 0.3 5.6 63144 14496 ? SN 04:53 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27942 1.4 5.6 63144 14496 ? SN 04:53 0:01 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27943 0.1 5.6 63144 14496 ? DN 04:53 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27944 0.2 5.6 63144 14496 ? SN 04:53 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27951 0.5 5.6 63144 14496 ? DN 04:54 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27955 0.0 5.6 63144 14496 ? SN 04:54 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27957 0.7 5.6 63144 14496 ? DN 04:54 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL

chrispadfield
Sun 3rd Jun '01, 10:21am
root 27965 0.0 5.6 63144 14496 ? DN 04:54 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27973 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27974 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27975 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27977 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27978 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27979 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27980 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27981 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27982 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27983 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27984 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
root 27987 0.0 5.6 63144 14496 ? SN 04:55 0:00 /usr/local/MySQL/libexec/mysqld --basedir=/usr/local/MySQL
nobody 27995 0.5 0.4 13976 1040 ? D 04:55 0:00 httpd
nobody 27997 2.6 0.5 2408 1448 ? D 04:55 0:00 perl ads.pl
root 27998 41.0 0.3 2548 900 pts/1 R 04:55 0:00 ps aux


but there is nothing that has been running for very long to cause this is there?
[/code]

chrispadfield
Sun 3rd Jun '01, 10:27am
i have also just read this at dialtoneinternet:


On many of our servers, by default, we mount your root partition with the 'sync' flag enabled. This disables buffered writing, which means the kernel immediately flushes your data to disk before returning success to your application.

This generally 'slows' down the application in question, but it does have the advantage of extra safety -- buffered writes are the most common cause of filesystem corruption in the case of a system crash or power failure. It's a fairly uncommon occurance in any case, but we prefer to keep the root partition -- the one with the absolute most important utilities -- extra safe.

However, for some clients, the slowdown is unacceptable, as it results in sometimes slow email handling, database performance (especially MySQL), and slow saving of configuration files on the root partition.

In order to disable sync writes, simply remove the 'sync' flag from /etc/fstab -- it will look somewhat like this:

/dev/sda1 / ext2 defaults,sync 1 1

Change it so that it looks like this:

/dev/sda1 / ext2 defaults 1 1

BE CAREFUL. Giving bad arguments to fstab can prevent your server from booting correctly the next time you reboot.

Once you've done this, you can enable async writes without needing to reboot. Simply execute:

mount -o remount,async FILESYSTEM

Where 'FILESYSTEM' is the mount point. For example:

mount -o remount,async /

You can see 'mount points' by executing the mount program with no arguments.

fastforward
Sun 3rd Jun '01, 2:51pm
Nice tip Chris. I like anthing that gives me extra speed... whatever the safety implications are :D

chrispadfield
Sun 3rd Jun '01, 7:12pm
well i emailed them about this and they said:


The problem is that your RAM is all used up and you are digging heavily into swap. This makes your disk 'thrash' swapping memory back and forth to your swapfile, and slows down your server. You need to double your RAM as soon as possible. Contact your account executive first thing Monday.

About the 'authorizing charges' thing: we don't keep track of your two free hours per month, the billing dept. does. When you need service, authorize charges as normal, and the billing dept. will deduct the charge from your free hours before charging your credit card for any remainder.

But in this case there's nothing technical to be done for you, you need to add RAM. I wish you had contacted us earlier about this, so you and your users wouldn't have had to suffer with a slow and crashing server for so long.
johncomeau 03-Jun-2001 16:01:02

We will continue to keep you informed as we work to complete your request.
This was in regard to your ticket:
>High load, low CPU usage

Thanks,
John Comeau


what do you think. 256MB ram and all there is is 25 people online, the phpads and also webadverts that does about 6,000 banners per day (banner exchange) apart from some email and some very light traffic sites that is all the server is being used for.

fastforward
Sun 3rd Jun '01, 7:31pm
Sounds a bit iffy. I've heard of more loaded servers running quite happily with just 128MB.

How much free space is on your disk? Their diagnosis may be correct, but there has to be a reason why this swapping is happening... and it can't be solely due to the main apps you have running.

chrispadfield
Sun 3rd Jun '01, 7:54pm
Filesystem Size Used Avail Use% Mounted on
/dev/hda1 102M 49M 48M 51% /
/dev/hda6 1.0G 832M 147M 85% /usr
/dev/hda7 1.0G 562M 418M 57% /var
/dev/hda8 6.1G 2.6G 3.1G 46% /home


and i agree completly iffy. Two virtual accounts would give me better performance. Very strange.

chrispadfield
Sun 3rd Jun '01, 8:03pm
i am wondering if this is just some cheap hard drive that is causing the problem. a SCSI hard drive sure would be better. If only i had some spare cash to move to venturesonline, am getting more and more tempted.

roy7
Mon 4th Jun '01, 12:57am
Another bit of info for you, load is based on processes in a run state waiting for something before they can execute. This delay can be needing CPU, disk accesses, network I/O, whatever. So if you have a load of 10 and cpu is 90% idle, that's a good indication something else is the bottleneck. Tools like iostat, vmstat, etc can help find them but it's a more complex issue than I can get into and not really my field anyway. :)

Linux doesn't sync writes by default, where as FreeBSD does. Linux did it for more performance, at the cost of reliability if the machine locks up or loses power. The way to get the best of both worlds is to have a caching controller (like a RAID controller has 32-128 meg of ram cache), with a battery backup. Have the OS sync, but the cache on the controller not sync directly to disk. If there's a power failure, those writes are still there and will be done when the power is restored. Thus, the OS is never fooled that a write happened which never goes to disk..

If you have a disk io bottleneck, also try splitting stuff between sperate drives.

If you have too little RAM, mysql has some suggestions in the manual about how to configure it to use less ram. But 256M is quite a bit, with only 20 users on the board, well, it should be fine. That's kinda odd.

-Jonathan

chrispadfield
Mon 4th Jun '01, 8:29am
thank you, i am learning a lot anyway. They support at dialtone, imbetween trying to charge you for pretty much anything, are just saying that "you need more ram".

Can someone explain how this can be the case. All the ram is being used by processes that arn't in use, ie are waiting. If there is not enough ram why don't these processes just go away so the ones running have enough ram to work. Seems bizarre to me.

roy7
Mon 4th Jun '01, 12:00pm
Processes not in use are swapped to disk and shouldn't use real ram.

Do something like "top -o res", which will sort by the Resident Memory field. Size is the amount of total memory allocated (which may or may not be shared libraries or shared with other processes). RES is the actualy amount of RAM the process is taking up. Typically with shared memory/etc, I see my httpd processes with a size around 5M-6M but a res of around 2M. For a forum with 200-300 users that I worked on, they will typically have about 150 http processes live at any time, which needs 300M-400M of ram.

Having tons of programs not doing anything shouldn't hurt since they would get swapped to disk, unless they are forcing themselves to remain in ram, but that's rare. Some OSes, like FreeBSD I think, won't swap it out unless the ram is needed though. I have some fairly idle processes that have ram still under RES, but I have free ram too.

-Jonathan

chrispadfield
Mon 4th Jun '01, 12:14pm
um... might be getting somewhere.

each httpd process is taking up 10 or 11 MB RAM! each mysql process around 9MB ram.

um...........

chrispadfield
Mon 4th Jun '01, 12:16pm
par example



PID USER PRI NI SIZE RSS SHARE STAT %CPU LIB %MEM TIME COMMAND
30093 nobody 9 0 11408 10M 1596 S 4.7 0 4.2 0:16 httpd
32612 root 17 0 604 604 300 R 4.5 0 0.2 0:05 top
32767 root 15 5 15256 8736 876 D N 3.6 0 3.3 0:05 mysqld
29462 nobody 6 0 11364 10M 1576 S 3.2 0 4.2 0:21 httpd
31569 nobody 6 0 11068 10M 1400 S 3.2 0 4.1 0:07 httpd
29777 nobody 5 0 10584 9.9M 748 S 2.8 0 3.9 0:15 httpd
32331 nobody 8 0 37256 18M 224 S 2.0 0 7.5 0:05 httpd
30235 root 7 5 15256 8736 876 S N 0.5 0 3.3 0:03 mysqld
5 root 3 0 0 0 0 DW 0.3 0 0.0 0:16 kswapd
28643 root 7 5 15256 8736 876 D N 0.3 0 3.3 0:07 mysqld
29510 root 6 5 15256 8736 876 S N 0.3 0 3.3 0:05 mysqld
401 named 0 0 2216 1648 376 S 0.1 0 0.6 0:17 named
30088 nobody 1 0 10768 10M 952 S 0.1 0 3.9 0:18 httpd
30447 nobody 0 0 10800 10M 952 S 0.1 0 4.0 0:14 httpd
31711 root 6 5 15256 8736 876 S N 0.1 0 3.3 0:04 mysqld
1 root 0 0 108 48 40 S 0.0 0 0.0 0:05 init

roy7
Mon 4th Jun '01, 12:16pm
Does the httpd have mod_perl? Even with all the stuff I add I never go over 6M resident for my httpd. mod_perl can add 4-5 megs per httpd, and it's one reason I dislike mod_perl. I greatly prefer fastcgi for faster perl cgi.

-Jonathan

chrispadfield
Mon 4th Jun '01, 12:17pm
no mod perl, i never installed it and seriously doubt it comes as standard.

roy7
Mon 4th Jun '01, 12:20pm
Hmmm. What's your total ram/etc lines from the top of top?

Maybe you have some php programs that use gobs of ram. I don't know. :)

-Jonathan

chrispadfield
Mon 4th Jun '01, 12:22pm
7:00am up 21:25, 1 user, load average: 0.23, 0.72, 0.71
213 processes: 210 sleeping, 3 running, 0 zombie, 0 stopped
CPU states: 17.6% user, 2.5% system, 3.5% nice, 76.2% idle
Mem: 257728K av, 253532K used, 4196K free, 53152K shrd, 5328K buff
Swap: 102808K av, 27988K used, 74820K free 32904K cached


i wonder if this all has something to do with my number of forums. I have an awful lot, due mainly to the books and films forums. Perhaps this causes the problem.

roy7
Mon 4th Jun '01, 12:28pm
If you are on a 64bit machine (Sun), it'll also use more ram than a typical 32 bit machine (Intel) would.

-Jonathan

chrispadfield
Mon 4th Jun '01, 12:33pm
nah, it is intel :)

chrispadfield
Mon 4th Jun '01, 12:35pm
thing is, i am not convinced it is just slow php scripts, because the mysql processes are also using up 9MB as well. Seems really strange.

chrispadfield
Mon 4th Jun '01, 12:36pm
sorry one more thing, is there a way to determine was page a particular httpd process is using, like if it is /home/ascifi/www/forums/forumdisplay.php sort of thing?

roy7
Mon 4th Jun '01, 12:39pm
If you have mod_status installed you can do something like yoursite.com/server-status to see a list of all connections and files being served, etc.

I've never used mysql on a machine that reports each thread individually, so I can't comment on it's usage. I'm used to seeing the entire memory used by all mysql threads as a single number. :) For big sites I might see a size of 300M with a res of 30M-50M.

-Jonathan

chrispadfield
Mon 4th Jun '01, 1:13pm
no mod_status.

so what do you think i should do? i guess upgrading php from 4.0.2 might help as well.

roy7
Mon 4th Jun '01, 1:32pm
Won't hurt, and if you use gzip in php, before 4.0.5 there was a memory leak. I prefer mod_gzip though. You could compile mod_status in. It isn't on by default in the basic apache config.

-Jonathan

chrispadfield
Mon 4th Jun '01, 2:50pm
this was here in the phpinfo thing


HTTP_SERVER_VARS["HTTP_ACCEPT_ENCODING"] gzip, deflate


and i have folders /bin/gzip and /usr/bin/gzip so i guess it is installed.

roy7
Mon 4th Jun '01, 2:53pm
No, /bin/gzip is just the gzip program installed on your machine. Accept encoding means your web browser can accept gzip'ed content. ;)

You'll almost certainly need to compile and install mod_gzip yourself, but you can use the built in php stuff too, it's safe in 4.0.5.

-Jonathan

chrispadfield
Mon 4th Jun '01, 3:04pm
ok, just wondering that was the possible source of a memory leak. There must be something weird going on for all the processes to be so high me thinks. I just wonder if it is all the forum I have. Deleting them will be the next step which would be a shame.

chrispadfield
Mon 4th Jun '01, 3:11pm
found something weird:


9:48am up 1 day, 13 min, 2 users, load average: 0.71, 1.01, 1.78
219 processes: 218 sleeping, 1 running, 0 zombie, 0 stopped
CPU states: 0.9% user, 2.3% system, 0.0% nice, 96.6% idle
Mem: 257728K av, 249392K used, 8336K free, 109068K shrd, 5688K buff
Swap: 102808K av, 19044K used, 83764K free 56068K cached

PID USER PRI NI SIZE RSS SHARE STAT LIB %CPU %MEM TIME COMMAND
7342 root 16 0 1140 1140 824 R 0 1.9 0.4 0:17 top
8000 nobody 1 0 13772 13M 2680 S 0 0.1 5.3 0:02 httpd
8038 nobody 14 0 5412 5412 2668 S 0 0.1 2.0 0:00 httpd
1 root 0 0 124 68 56 S 0 0.0 0.0 0:05 init
2 root 0 0 0 0 0 SW 0 0.0 0.0 0:00 kflushd
3 root 0 0 0 0 0 SW 0 0.0 0.0 0:00 kupdate
4 root 0 0 0 0 0 SW 0 0.0 0.0 0:00 kpiod
5 root 0 0 0 0 0 SW 0 0.0 0.0 0:26 kswapd
6 root -20 -20 0 0 0 SW< 0 0.0 0.0 0:00 mdrecoveryd
273 root 0 0 76 0 0 SW 0 0.0 0.0 0:00 apmd
326 root 0 0 212 152 124 S 0 0.0 0.0 0:01 syslogd
337 root 0 0 372 0 0 SW 0 0.0 0.0 0:00 klogd
353 daemon 0 0 144 104 76 S 0 0.0 0.0 0:00 atd
369 root 1 0 164 108 80 S 0 0.0 0.0 0:00 crond
385 root 0 0 72 0 0 SW 0 0.0 0.0 0:00 inetd
401 named 0 0 2608 2184 648 S 0 0.0 0.8 0:22 named
417 root 0 0 84 0 0 SW 0 0.0 0.0 0:00 lpd
427 qmails 0 0 184 152 104 S 0 0.0 0.0 0:08 qmail-send
428 qmaill 0 0 80 20 12 S 0 0.0 0.0 0:00 cyclog
429 root 0 0 112 60 48 S 0 0.0 0.0 0:00 tcpserver
430 qmails 0 0 112 60 48 S 0 0.0 0.0 0:00 tcpserver
437 qmaill 0 0 76 24 12 S 0 0.0 0.0 0:01 accustamp
438 root 0 0 116 72 56 S 0 0.0 0.0 0:00 qmail-lspawn
439 qmailr 1 0 136 100 84 S 0 0.0 0.0 0:00 qmail-rspawn
440 qmailq 0 0 124 92 68 S 0 0.0 0.0 0:00 qmail-clean
757 root 0 0 1368 44 28 S 0 0.0 0.0 0:01 httpsd
760 nobody 0 0 180 0 0 SW 0 0.0 0.0 0:00 gcache
776 root 0 0 68 0 0 SW 0 0.0 0.0 0:00 mingetty
777 root 0 0 68 0 0 SW 0 0.0 0.0 0:00 mingetty
778 root 0 0 84 0 0 SW 0 0.0 0.0 0:00 getty
781 nobody 0 0 1424 0 0 SW 0 0.0 0.0 0:00 httpsd
782 nobody 0 0 1604 0 0 SW 0 0.0 0.0 0:00 httpsd
783 nobody 0 0 1360 0 0 SW 0 0.0 0.0 0:00 httpsd
784 nobody 0 0 1360 0 0 SW 0 0.0 0.0 0:00 httpsd
785 nobody 0 0 1592 0 0 SW 0 0.0 0.0 0:00 httpsd
4153 admin 0 0 220 0 0 SW 0 0.0 0.0 0:00 bash
4164 root 0 0 360 0 0 SW 0 0.0 0.0 0:00 su
4165 root 0 0 216 0 0 SW 0 0.0 0.0 0:00 bash
4174 root 0 0 144 0 0 SW 0 0.0 0.0 0:00 safe_mysqld
4185 root 5 5 16276 10M 4236 S N 0 0.0 4.1 0:01 mysqld
4187 root 16 5 16276 10M 4236 S N 0 0.0 4.1 0:02 mysqld
4188 root 5 5 16276 10M 4236 S N 0 0.0 4.1 0:00 mysqld
8876 nobody 0 0 1364 0 0 SW 0 0.0 0.0 0:00 httpsd

chrispadfield
Mon 4th Jun '01, 3:12pm
is this right:


4185 root 5 5 16276 10M 4236 S N 0 0.0 4.1 0:01 mysqld
4187 root 16 5 16276 10M 4236 S N 0 0.0 4.1 0:02 mysqld
4188 root 5 5 16276 10M 4236 S N 0 0.0 4.1 0:00 mysqld


taking up the RSS but not being used at that time (ie not cpu) ?

chrispadfield
Mon 4th Jun '01, 3:14pm
and this is it sorted by memory use


8000 nobody 1 0 12856 12M 1540 S 0 0.0 4.9 0:02 httpd
8002 nobody 1 0 11804 11M 2100 S 0 0.0 4.4 0:03 httpd
8013 nobody 13 0 11708 11M 2204 S 0 11.3 4.4 0:03 httpd
8007 nobody 1 0 11596 11M 1992 S 0 0.0 4.4 0:02 httpd
8003 nobody 0 0 11564 11M 1980 S 0 0.0 4.4 0:02 httpd
7996 nobody 3 0 11560 11M 1896 S 0 0.0 4.4 0:03 httpd
8066 nobody 0 0 11432 11M 2044 S 0 0.0 4.3 0:01 httpd
7998 nobody 0 0 11340 10M 2052 S 0 0.0 4.3 0:01 httpd
8241 nobody 1 0 11360 10M 2192 S 0 0.1 4.3 0:00 httpd
8068 nobody 0 0 11280 10M 2072 S 0 0.0 4.2 0:01 httpd
8035 nobody 1 0 11080 10M 1780 S 0 0.1 4.2 0:02 httpd
8039 nobody 0 0 11112 10M 1396 S 0 0.0 4.2 0:02 httpd
8015 nobody 0 0 10788 10M 1388 S 0 0.0 4.1 0:01 httpd
7997 nobody 0 0 10716 10M 1344 S 0 0.0 4.0 0:01 httpd
8014 nobody 0 0 10716 10M 1388 S 0 0.0 4.0 0:00 httpd
8040 nobody 0 0 10620 10M 1412 S 0 0.0 4.0 0:00 httpd
4185 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:01 mysqld
4187 root 8 5 16052 8064 2268 S N 0 0.0 3.1 0:02 mysqld
4188 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8008 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:01 mysqld
8022 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8028 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8036 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8041 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8056 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8063 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8084 root 6 5 16052 8064 2268 S N 0 0.9 3.1 0:02 mysqld
8094 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8110 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:01 mysqld
8113 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8117 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8119 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8124 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:01 mysqld
8151 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8153 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8171 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8205 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8207 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8222 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8250 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8281 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8283 root 5 5 16052 8064 2268 S N 0 0.0 3.1 0:00 mysqld
8009 nobody 0 0 5008 4876 2232 S 0 0.0 1.8 0:00 httpd
8038 nobody 2 0 4928 4720 2008 S 0 0.0 1.8 0:00 httpd
8016 nobody 1 0 4808 4596 1824 S 0 0.0 1.7 0:01 httpd
8018 nobody 0 0 4672 4472 2000 S 0 0.0 1.7 0:00 httpd
8020 nobody 0 0 4568 4360 1524 S 0 0.0 1.6 0:02 httpd
8001 nobody 0 0 3444 3204 388 S 0 0.0 1.2 0:01 httpd

Dave#
Mon 4th Jun '01, 8:11pm
Chris,

I'm with DI too and I have a few problems with my serverload :)

Anyways a couple of things:

What version of Mysql are you running? The latest builds have a remarkable effect.

I would suggest upgrading PHP it may help a little.

The EIDE disks that DI ships as standard are definately a bottleneck


The other things I have done is

Got some excellent Mysql tuning tips in this forum
Disabled all APACHE logging apart from error logging

After all that I can run 50-200 concurrent users on my p2400 with 512 meg of ram and eide disks (although I am considering getting another disk for Mysql only)

<Secret DI tip: Get a memory upgrade somepeople report that DI don't charge the monthly fee for the upgrade :) >

fastforward
Mon 4th Jun '01, 9:19pm
Originally posted by Dave#
<Secret DI tip: Get a memory upgrade somepeople report that DI don't charge the monthly fee for the upgrade :) >
I just got my 256MB memory upgrade instaled today. $185 once only fee. That brings me up to 512MB now :)

Also, I'm sure you know, but DI now ship SCSI as standard. Mine came with a 10Krpm 9.1GB. Also, they advertise a 650mhz CPU but mine came with a 750. (you think I should complain and send it back? :D )

Is it possible for you to simply cancel and re-order one of their later servers? There's no setup fee and you've seen your 6 month contract through.

chrispadfield
Mon 4th Jun '01, 10:25pm
no, i already upgrade to 128 from 256 previously and that extended the contract.

The other thing is i had someone install qmail for me, and that was not cheap. I couldn't do sales@domain1.com and sales@domain2.com as seperate pop3 accounts otherwise which is one of the things i needed the server for.

If i do change servers, i will go for VO mainly because of cpanel and webhostmanager that will allow me to easily resale some of the webspace.

Dave#
Tue 5th Jun '01, 5:49pm
If i do change servers, i will go for VO mainly because of cpanel and webhostmanager that will allow me to easily resale some of the webspace.

You may be in for a shock when you go for a quote :(

chrispadfield
Tue 5th Jun '01, 8:14pm
how do you mean? It is $400/month for the PIII 866 with 256MB and 18GB SCSI but more importantly with 100GB of transfer. My dialtone one is something like $329/month for PIII 650 with 256MB RAM, no scsi hard drive and only 65GB transfer. The extra 35GB * $4 (dialtone rate) is easily more than the difference. Plus, $50 of VO goes towards the payment for cpanel, adding extra ram/hard drive/processors is cheap and mysql and php are supported products and the tech support rocks, and they have their own forum etc...

Thing i like about dialtone is they aren't going to go anywhere, VO could and the dialtone data center has been incredibly reliable whereas even the verio one is not as reliable as one would think and the less said about vdi the better.

In the end, just being able to sell some webspace of and control it is what i want. I am not linux admin but learning a lot. cpanel although not brilliant by any stretch of the imagination should help.

fastforward
Tue 5th Jun '01, 8:37pm
Go on then... sod off to VO! See if we care :D

Nah, you make a good case for moving. I guess it depends what you want the server for. For me, Dialtone is perfect. Reliable fast, I don't need support so it's cheap. Plus there aren't many servers available with SCSI for this price. I'm paying $300 for 750Mhz, 9.1GB SCSI and 65GB transfer. I do have one complaint and that is the O/S version. I wish they had an option for FreeBSD or at least Linux 6.2.

But I won't be moving anywhere in a hurry (unless they start double billing me). I've heard a few horror stories about their accounting department. In fact, just last week I changed my billing method from credit card to cheque just in case :).