Samples of TOP command


#1

Can I see samples of TOP results from another typical user?

Mine is strange…

Is it normal to have 40 sleeping process? I mean if I see my sites, my sites are generated in less than 1 seconds. After that I expect memory to be released… But no… There are so many sleeping processes that’s running for a long time…

What’s going on? Is there a way that after my program finishes generating site and my site response to the get command from the other server, that it just quit and releases memory?

Can I show others’ user typical top request?

I think that’s the main problem… Somehow it takes a while for memories to get released even though my program has finished a very long time…

That being said, I don’t think that kind of memory will affect server’s performance… SWAP memories are cheap and only take hard disk space… Most of the time, those sleeping processes only take on swap memories…

I do recommend that you increase the ratio between SWAP and real memories.

Teguh

Here is my TOP request:
top - 03:53:41 up 3 days, 3:59, 2 users, load average: 0.26, 1.22, 1.74
Tasks: 40 total, 3 running, 34 sleeping, 0 stopped, 3 zombie
Cpu(s): 10.4% us, 3.7% sy, 0.0% ni, 85.6% id, 0.0% wa, 0.0% hi, 0.3% si
Mem: 417792k total, 340764k used, 77028k free, 0k buffers
Swap: 417792k total, 0k used, 417792k free, 0k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ PPID SWAP TIME CODE DATA COMMAND
26280 teguh123 16 0 0 0 0 Z 1.7 0.0 0:00.05 23805 0 0:00 0 0 php5.cgi
26282 teguh123 17 0 20568 8808 4680 R 1.7 2.1 0:00.05 26281 11m 0:00 6504 3492 php5.cgi
26270 teguh123 15 0 20364 8912 4920 S 1.3 2.1 0:00.04 22845 11m 0:00 6504 3232 php5.cgi
26274 obd 16 0 0 0 0 Z 1.3 0.0 0:00.04 23507 0 0:00 0 0 php5.cgi
26276 teguh123 16 0 0 0 0 Z 1.3 0.0 0:00.04 24293 0 0:00 0 0 php5.cgi
22845 dhapache 15 0 31864 18m 3864 S 0.3 4.6 0:00.35 2445 12m 0:00 316 15m apache2-teguh12
22854 dhapache 15 0 32048 18m 3868 S 0.3 4.6 0:00.29 2445 12m 0:00 316 15m apache2-teguh12
24101 dhapache 15 0 31968 18m 3864 S 0.3 4.6 0:00.21 2445 12m 0:00 316 15m apache2-teguh12
24293 dhapache 15 0 32048 18m 3864 S 0.3 4.6 0:00.18 2445 12m 0:00 316 15m apache2-teguh12
24299 dhapache 15 0 31968 18m 3864 S 0.3 4.6 0:00.15 2445 12m 0:00 316 15m apache2-teguh12
25849 dhapache 15 0 31904 18m 3860 S 0.3 4.6 0:00.05 2445 12m 0:00 316 15m apache2-teguh12
1 root 15 0 1848 576 496 S 0.0 0.1 0:01.85 0 1272 0:01 32 280 init
2407 root 22 0 2820 1512 676 S 0.0 0.4 0:15.56 1 1308 0:15 276 960 syslog-ng
2445 root 15 0 31468 20m 5840 S 0.0 4.9 0:23.31 1 10m 0:23 316 14m apache2-teguh12
2461 root 15 0 1616 480 420 S 0.0 0.1 0:00.03 1 1136 0:00 20 232 inetd
2572 root 23 0 3044 1156 980 S 0.0 0.3 0:01.22 1 1888 0:01 24 264 master
2577 postfix 17 0 3084 1228 1016 S 0.0 0.3 0:01.13 2572 1856 0:01 40 264 qmgr
2582 root 15 0 3548 928 696 S 0.0 0.2 0:00.63 1 2620 0:00 276 456 sshd
2593 nobody 18 0 2464 832 472 S 0.0 0.2 0:00.14 1 1632 0:00 428 580 proftpd
2596 daemon 18 0 1768 388 308 S 0.0 0.1 0:00.00 1 1380 0:00 16 248 atd
2599 root 19 0 1824 756 608 S 0.0 0.2 0:03.53 1 1068 0:03 28 248 cron
2609 root 18 0 1572 448 388 S 0.0 0.1 0:00.04 1 1124 0:00 16 232 ud
9612 teguh123 15 0 2132 1068 836 S 0.0 0.3 0:04.20 31248 1064 0:04 48 280 top
16508 dhapache 24 0 31468 15m 1196 S 0.0 3.8 0:00.00 2445 15m 0:00 316 14m apache2-teguh12
23507 dhapache 15 0 31880 18m 3864 S 0.0 4.6 0:00.23 2445 12m 0:00 316 15m apache2-teguh12
23634 dhapache 15 0 31960 18m 3864 S 0.0 4.6 0:00.25 2445 12m 0:00 316 15m apache2-teguh12
23805 dhapache 15 0 31992 18m 3864 S 0.0 4.6 0:00.25 2445 12m 0:00 316 15m apache2-teguh12
24282 dhapache 15 0 31860 18m 4024 S 0.0 4.6 0:00.18 2445 12m 0:00 316 15m apache2-teguh12
24296 dhapache 15 0 31864 18m 3864 S 0.0 4.6 0:00.16 2445 12m 0:00 316 15m apache2-teguh12
25812 root 15 0 6124 1540 1244 S 0.0 0.4 0:00.00 2582 4584 0:00 276 456 sshd
25848 dhapache 15 0 31920 18m 3864 S 0.0 4.6 0:00.03 2445 12m 0:00 316 15m apache2-teguh12
25850 dhapache 15 0 32008 18m 3864 S 0.0 4.6 0:00.06 2445 12m 0:00 316 15m apache2-teguh12
25866 teguh123 15 0 6108 1216 912 R 0.0 0.3 0:00.00 25812 4892 0:00 276 456 sshd
25871 teguh123 15 0 2608 1384 1128 S 0.0 0.3 0:00.00 25866 1224 0:00 592 228 bash
25890 teguh123 15 0 2132 1040 836 R 0.0 0.2 0:00.00 25871 1092 0:00 48 280 top
26281 dhapache 16 0 31736 18m 3768 S 0.0 4.5 0:00.00 2445 12m 0:00 316 15m apache2-teguh12
29694 postfix 19 0 3052 1172 992 S 0.0 0.3 0:00.00 2572 1880 0:00 8 264 pickup
31130 root 15 0 6124 1536 1240 S 0.0 0.4 0:00.00 2582 4588 0:00 276 456 sshd
31243 teguh123 15 0 6108 1220 912 S 0.0 0.3 0:01.60 31130 4888 0:01 276 456 sshd
31248 teguh123 15 0 2608 1384 1128 S 0.0 0.3 0:00.00 31243 1224 0:00 592 228 bash


#2

Please don’t post your request for help with ‘top’ multiple times and in multiple forums.

-Scott


#3

Look, this is a different question… I was asking a sample of TOP commands that I would like to compare with my own TOP command.


#4

Look, You can find all the “samples” of “top commands” you want in the top man pages, and there is no need to “compare” them with anything because your “own TOP command” is no different than mine or anyone else’s using the same version of top.

–rlparker


#5

I am sorry… I mean samples of TOP results:

Here is my TOP request:
top - 03:53:41 up 3 days, 3:59, 2 users, load average: 0.26, 1.22, 1.74
Tasks: 40 total, 3 running, 34 sleeping, 0 stopped, 3 zombie
Cpu(s): 10.4% us, 3.7% sy, 0.0% ni, 85.6% id, 0.0% wa, 0.0% hi, 0.3% si
Mem: 417792k total, 340764k used, 77028k free, 0k buffers
Swap: 417792k total, 0k used, 417792k free, 0k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ PPID SWAP TIME CODE DATA COMMAND
26280 teguh123 16 0 0 0 0 Z 1.7 0.0 0:00.05 23805 0 0:00 0 0 php5.cgi
26282 teguh123 17 0 20568 8808 4680 R 1.7 2.1 0:00.05 26281 11m 0:00 6504 3492 php5.cgi
26270 teguh123 15 0 20364 8912 4920 S 1.3 2.1 0:00.04 22845 11m 0:00 6504 3232 php5.cgi
26274 obd 16 0 0 0 0 Z 1.3 0.0 0:00.04 23507 0 0:00 0 0 php5.cgi
26276 teguh123 16 0 0 0 0 Z 1.3 0.0 0:00.04 24293 0 0:00 0 0 php5.cgi
22845 dhapache 15 0 31864 18m 3864 S 0.3 4.6 0:00.35 2445 12m 0:00 316 15m apache2-teguh12
22854 dhapache 15 0 32048 18m 3868 S 0.3 4.6 0:00.29 2445 12m 0:00 316 15m apache2-teguh12
24101 dhapache 15 0 31968 18m 3864 S 0.3 4.6 0:00.21 2445 12m 0:00 316 15m apache2-teguh12
24293 dhapache 15 0 32048 18m 3864 S 0.3 4.6 0:00.18 2445 12m 0:00 316 15m apache2-teguh12
24299 dhapache 15 0 31968 18m 3864 S 0.3 4.6 0:00.15 2445 12m 0:00 316 15m apache2-teguh12
25849 dhapache 15 0 31904 18m 3860 S 0.3 4.6 0:00.05 2445 12m 0:00 316 15m apache2-teguh12
1 root 15 0 1848 576 496 S 0.0 0.1 0:01.85 0 1272 0:01 32 280 init
2407 root 22 0 2820 1512 676 S 0.0 0.4 0:15.56 1 1308 0:15 276 960 syslog-ng
2445 root 15 0 31468 20m 5840 S 0.0 4.9 0:23.31 1 10m 0:23 316 14m apache2-teguh12
2461 root 15 0 1616 480 420 S 0.0 0.1 0:00.03 1 1136 0:00 20 232 inetd
2572 root 23 0 3044 1156 980 S 0.0 0.3 0:01.22 1 1888 0:01 24 264 master
2577 postfix 17 0 3084 1228 1016 S 0.0 0.3 0:01.13 2572 1856 0:01 40 264 qmgr
2582 root 15 0 3548 928 696 S 0.0 0.2 0:00.63 1 2620 0:00 276 456 sshd
2593 nobody 18 0 2464 832 472 S 0.0 0.2 0:00.14 1 1632 0:00 428 580 proftpd
2596 daemon 18 0 1768 388 308 S 0.0 0.1 0:00.00 1 1380 0:00 16 248 atd
2599 root 19 0 1824 756 608 S 0.0 0.2 0:03.53 1 1068 0:03 28 248 cron
2609 root 18 0 1572 448 388 S 0.0 0.1 0:00.04 1 1124 0:00 16 232 ud
9612 teguh123 15 0 2132 1068 836 S 0.0 0.3 0:04.20 31248 1064 0:04 48 280 top
16508 dhapache 24 0 31468 15m 1196 S 0.0 3.8 0:00.00 2445 15m 0:00 316 14m apache2-teguh12
23507 dhapache 15 0 31880 18m 3864 S 0.0 4.6 0:00.23 2445 12m 0:00 316 15m apache2-teguh12
23634 dhapache 15 0 31960 18m 3864 S 0.0 4.6 0:00.25 2445 12m 0:00 316 15m apache2-teguh12
23805 dhapache 15 0 31992 18m 3864 S 0.0 4.6 0:00.25 2445 12m 0:00 316 15m apache2-teguh12
24282 dhapache 15 0 31860 18m 4024 S 0.0 4.6 0:00.18 2445 12m 0:00 316 15m apache2-teguh12
24296 dhapache 15 0 31864 18m 3864 S 0.0 4.6 0:00.16 2445 12m 0:00 316 15m apache2-teguh12
25812 root 15 0 6124 1540 1244 S 0.0 0.4 0:00.00 2582 4584 0:00 276 456 sshd
25848 dhapache 15 0 31920 18m 3864 S 0.0 4.6 0:00.03 2445 12m 0:00 316 15m apache2-teguh12
25850 dhapache 15 0 32008 18m 3864 S 0.0 4.6 0:00.06 2445 12m 0:00 316 15m apache2-teguh12
25866 teguh123 15 0 6108 1216 912 R 0.0 0.3 0:00.00 25812 4892 0:00 276 456 sshd
25871 teguh123 15 0 2608 1384 1128 S 0.0 0.3 0:00.00 25866 1224 0:00 592 228 bash
25890 teguh123 15 0 2132 1040 836 R 0.0 0.2 0:00.00 25871 1092 0:00 48 280 top
26281 dhapache 16 0 31736 18m 3768 S 0.0 4.5 0:00.00 2445 12m 0:00 316 15m apache2-teguh12
29694 postfix 19 0 3052 1172 992 S 0.0 0.3 0:00.00 2572 1880 0:00 8 264 pickup
31130 root 15 0 6124 1536 1240 S 0.0 0.4 0:00.00 2582 4588 0:00 276 456 sshd
31243 teguh123 15 0 6108 1220 912 S 0.0 0.3 0:01.60 31130 4888 0:01 276 456 sshd
31248 teguh123 15 0 2608 1384 1128 S 0.0 0.3 0:00.00 31243 1224 0:00 592 228 bash

I have so many sleeping processes that’s idle and taking up memory… My pages are generated within less than 1 seconds but each of those sleeping processes take minutes to die

16508 dhapache 24 0 31468 15m 1196 S 0.0 3.8 0:00.00 2445 15m 0:00 316 14m apache2-teguh12
23507 dhapache 15 0 31880 18m 3864 S 0.0 4.6 0:00.23 2445 12m 0:00 316 15m apache2-teguh12
23634 dhapache 15 0 31960 18m 3864 S 0.0 4.6 0:00.25 2445 12m 0:00 316 15m apache2-teguh12
23805 dhapache 15 0 31992 18m 3864 S 0.0 4.6 0:00.25 2445 12m 0:00 316 15m apache2-teguh12
24282 dhapache 15 0 31860 18m 4024 S 0.0 4.6 0:00.18 2445 12m 0:00 316 15m apache2-teguh12
24296 dhapache 15 0 31864 18m 3864 S 0.0 4.6 0:00.16 2445 12m 0:00 316 15m apache2-teguh12
25812 root 15 0 6124 1540 1244 S 0.0 0.4 0:00.00 2582 4584 0:00 276 456 sshd
25848 dhapache 15 0 31920 18m 3864 S 0.0 4.6 0:00.03 2445 12m 0:00 316 15m apache2-teguh12
25850 dhapache 15 0 32008 18m 3864 S 0.0 4.6 0:00.06 2445 12m 0:00 316 15m apache2-teguh12

I wonder if others are like me?


#6

When apache2 starts it creates multiple threads to handle future connections. In the above top, process 2445 (running as root) is the first call. root then creates the other apache2’s to handle potential connections and keeps some spare for new connections.

You might be able to change the number of retained processes in apache2.conf but this could be at the epxense of overall website functionality if your site is popular. Take a look in /usr/share/doc/apache2-doc/manual/index.html for info on apache2.

Maximum Cash Discount on any plan with MAXCASH


#7

You mean all those sleeping stuffs are not even handling an active connections but just reserved for future “potential” connections?

Those sleeping processes take so much memory… I am not allowed to move back to reseller account simply because of that despite the fact that my CPU takes only 10% of what it should be…

Time to re ask the support then…

How would my website functionality will be hurt if I greatly reduce the number of retained processes? You mean they’ll be slower?


#8

No, not all, but apache2 does have sleepers to handle new connections.

Perhaps you’re being overly concerned about the top results showing sleepers and memory used when it really isn’t a concern unless it’s actually impacting on your website’s functionality. Unused CPU and MEM on a computer is an indication of waste rather than anything imho.

The main “thing” is load - and the load you’ve reported above is awesome :wink:

Maximum Cash Discount on any plan with MAXCASH


#9

Well, dreamhost support keep insisting that unless I bring memory usage down I can’t go back to regular hosting…

From top, the one that uses the most memory are those sleeping process…

The load are awesome? What do you mean? Good or bad? Load is quite low with occasional spikes.

Here, I sort my top request based on username.

While there are occasional spikes in load, most of the time it’s low. Most of the memories are taken by sleeping process taking 15 M each… Multiplied that by 40 and I see where I have memory problems.

top - 22:36:37 up 5 days, 22:42, 1 user, load average: 0.35, 0.33, 0.51
Tasks: 41 total, 1 running, 40 sleeping, 0 stopped, 0 zombie
Cpu(s): 15.9% us, 3.3% sy, 0.0% ni, 77.0% id, 3.7% wa, 0.0% hi, 0.2% si
Mem: 417792k total, 417792k used, 0k free, 0k buffers
Swap: 417792k total, 19268k used, 398524k free, 0k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME SWAP CODE DATA WCHAN COMMAND
8365 teguh123 15 0 6108 1216 912 S 0.0 0.3 0:00.00 4892 276 456 sinittext sshd
8366 teguh123 15 0 2608 1384 1128 S 0.0 0.3 0:00.00 1224 592 228 wait bash
8989 teguh123 15 0 2132 1056 836 R 0.0 0.3 0:00.02 1076 48 280 sinittext top
9660 teguh123 15 0 20364 8888 4928 S 1.7 2.1 0:00.05 11m 6504 3232 sinittext php5.cgi
9663 teguh123 15 0 20884 9288 5004 S 2.3 2.2 0:00.07 11m 6504 3752 rpc_wait_ php5.cgi
1 root 15 0 1848 576 496 S 0.0 0.1 0:01.94 1272 32 280 sinittext init
2445 root 15 0 31584 20m 5840 S 0.0 5.0 0:47.74 10m 316 15m sinittext apache2-teguh12
2461 root 18 0 1616 480 420 S 0.0 0.1 0:00.03 1136 20 232 sinittext inetd
2572 root 15 0 3044 1156 980 S 0.0 0.3 0:02.42 1888 24 264 sinittext master
2582 root 18 0 3548 928 696 S 0.0 0.2 0:01.15 2620 276 456 sinittext sshd
2599 root 16 0 1824 756 608 S 0.0 0.2 0:06.88 1068 28 248 sinittext cron
2609 root 23 0 1572 468 408 S 0.0 0.1 0:01.15 1104 16 232 sinittext ud
8326 root 15 0 6124 1536 1240 S 0.0 0.4 0:00.00 4588 276 456 sinittext sshd
2577 postfix 15 0 3084 1228 1016 S 0.0 0.3 0:02.26 1856 40 264 sinittext qmgr
4597 postfix 15 0 3052 1172 992 S 0.0 0.3 0:00.00 1880 8 264 sinittext pickup
8932 postfix 16 0 3400 1660 1408 S 0.0 0.4 0:00.00 1740 28 436 sinittext cleanup
8933 postfix 16 0 3232 1456 1244 S 0.0 0.3 0:00.00 1776 16 436 sinittext trivial-rewrite
8937 postfix 16 0 3252 1532 1304 S 0.0 0.4 0:00.00 1720 36 436 sinittext local
8939 postfix 16 0 3108 1336 1132 S 0.0 0.3 0:00.00 1772 32 264 sinittext smtp
2593 nobody 18 0 2464 836 472 S 0.0 0.2 0:00.31 1628 428 580 sinittext proftpd
3037 dhapache 18 0 32064 18m 3868 S 0.0 4.6 0:00.30 12m 316 15m 377190947 apache2-teguh12
3365 dhapache 18 0 32048 18m 3868 S 0.0 4.6 0:00.28 12m 316 15m nfs_wait_ apache2-teguh12
7127 dhapache 18 0 31936 18m 3864 S 0.0 4.6 0:00.10 12m 316 15m nfs_wait_ apache2-teguh12
7128 dhapache 15 0 31936 18m 3868 S 0.0 4.6 0:00.13 12m 316 15m 339274961 apache2-teguh12
7129 dhapache 15 0 31936 18m 3864 S 0.0 4.6 0:00.09 12m 316 15m rpc_wait_ apache2-teguh12
7737 dhapache 15 0 32016 18m 3864 S 0.0 4.6 0:00.07 12m 316 15m 345817358 apache2-teguh12
8295 dhapache 15 0 31928 18m 3860 S 0.0 4.6 0:00.02 12m 316 15m sinittext apache2-teguh12
8316 dhapache 15 0 31920 18m 3864 S 0.0 4.6 0:00.03 12m 316 15m 335122846 apache2-teguh12
8331 dhapache 15 0 32032 18m 3892 S 0.0 4.6 0:00.04 12m 316 15m sinittext apache2-teguh12
9303 dhapache 15 0 32032 18m 3856 S 0.0 4.6 0:00.00 12m 316 15m sinittext apache2-teguh12
9394 dhapache 15 0 31912 18m 3856 S 0.0 4.6 0:00.01 12m 316 15m 374325928 apache2-teguh12
9397 dhapache 15 0 31936 18m 3884 S 0.0 4.6 0:00.03 12m 316 15m 344191800 apache2-teguh12
9398 dhapache 15 0 32040 18m 3864 S 0.0 4.6 0:00.01 12m 316 15m nfs_wait_ apache2-teguh12
9451 dhapache 18 0 31904 18m 3860 S 0.0 4.6 0:00.00 12m 316 15m nfs_wait_ apache2-teguh12
9452 dhapache 18 0 31984 18m 3892 S 0.3 4.6 0:00.02 12m 316 15m sinittext apache2-teguh12
9453 dhapache 16 0 31928 18m 3860 S 0.0 4.6 0:00.00 12m 316 15m 414797275 apache2-teguh12
9456 dhapache 15 0 31856 18m 3852 S 0.3 4.6 0:00.02 12m 316 15m 345130326 apache2-teguh12
9457 dhapache 15 0 31952 18m 3888 S 0.0 4.6 0:00.01 12m 316 15m sinittext apache2-teguh12
9458 dhapache 15 0 31920 18m 3884 S 0.0 4.6 0:00.00 12m 316 15m 369991540 apache2-teguh12
17807 dhapache 15 0 31584 15m 1212 S 0.0 3.9 0:00.00 15m 316 15m sinittext apache2-teguh12
2596 daemon 15 0 1768 388 308 S 0.0 0.1 0:00.00 1380 16 248 sinittext atd


#10

They’re really good! You won’t see results like those on a shared server.

I take it your site is functioning well.

Maximum Cash Discount on any plan with MAXCASH


#11

This is what support says again…

Your entire server without your processes uses around 30 MB or RAM, you’re pushing it up over 800 MB which is not going to fly under shared hosting (I am sorry but the processes you are running are not going to fit under shared hosting if you’re taking a quarter of the RAM OR CPU - likely both in this case). While we can allow some disproportionate usage at some point you have to pay for what you use and your account is above that threshold.

I can set up procwatch for you if you like to see if that helps (it will kill off your processes that are not releasing memory)… we can’t move you to shared until you can function with the slider all the way to the left.

Thanks!

John


#12

DreamHost PS tries to set the range of waiting apaches to use the available RAM so as to give you the best performance. Where apache settles in is dependent on your site’s traffic. If you move the slider down it will lower the range of waiting apaches. Have you tried lowering the slider? Does your website still function well?
At any given moment many of the apache processes might be waiting for incoming connections. But, there is overhead to dropping and restarting the processes, so you don’t want to always have to spawn a new one for each new request.
As John points out if you can move the slider all the way down and your site still works fine, then you didn’t need the extra RAM and apaches. If you notice a performance hit, then your site actually needs those resources to perform well.


#13

Most of the time, all the way left work. However, I increase the bar to ensure stability…

Also most of my sites are generated within less than 1 seconds. I instruct PhP to die after 8 seconds.

One thing for sure is I don’t think I need 60 sleeping Apache process to catch the ball to speed things up when someone visits. Most of the time, there are only 0 or 2 visitors that visit my sites at the same time. During very rare peak there could be 5 or 6. Why would I need 60 sleeping apache processes ready to catch balls?

Basically, CPU is no longer a problem. As for memory, you said those things won’t show up if I am no longer using PS?

Well… Why not try it for 2 days? See if I cause instability. That’s all I am asking.

I am just curious… If I order 150 MB memory from PS, I got 150 MB memory plus 150 MB swap memory…

That swap memory costs only hard disk space right? Why not give your customers 1.5 G of swap memory rather than 150MB? I thought hard disk space are cheap…


#14

Glad to hear that you were able to find a better price point that still performs well. I would place the slider to rarely ever need swap, since the performance hit for using it is so great.


#15

Correction again… As of now, my programs typically spend .008 seconds to generate a page… So yes it’s very efficient now. Sometimes, during peak, it’ll take 1 seconds… But that’s rare.


What do you mean by this: “I would place the slider to rarely ever need swap, since the performance hit for using it is so great.”

I thought Swap memory was great, especially if all it does is baby siting sleeping process anyway.

I think my biggest problem is that I have a very efficient site that take less than 1 seconds for each call. However, apache somehow overestimate the amount of sleeping processes assigned for me.

I don’t think this will be a problem in a normal server. As you see I am not using many of those sleeping processes anyway. I’ll be glad to lower those sleeping processes to a maximum of 6 or something.

I want to try normal server again for 2 days. The server was stable when I moved to PS. At that time, the issue was CPU power because I called gethostbyaddress for every visit.

Now, with that being fixed plus some more fine tuning, CPU is no longer the issue.

I don’t think I’ll cause instability if I move back to normal server. Please give me 1 or 2 days try. If there is a problem, I’ll move back up to PS quickly.


#16

I have not examined the usage patterns for your website, so I can not speak to those. Swap is not good since it is information written to the disk rather than kept in memory and memory access is much faster than disk access.

If you want to try and fit your site into shared hosting which offers less resources than a PS, you should leave the slider all the way down and show john that your memory and cpu are not clipping at that level.

My guess would be that you are mistaken about not needing the resources if you can’t leave the slider all the way to the left at all times without performance degradation, since that would be the definition of not needing it.

If your website is not generating enough money to cover its resource usage, from a business standpoint I would suggest optimizing the website code, caching more content, or perhaps using static content where possible. If its just a personal site, I would suggest removing any resource intensive functionality that doesn’t give you more joy than the cost of providing that functionality. Perhaps there is a badly coded plugin that you could disable, or perhaps there is some comment spam that you could block.

Where is most of your usage coming from? Are there a couple ips which account for most of your traffic but don’t buy anything or add new content to your site? Perhaps there is a search engine crawler that is crawling your entire website all the time without much benefit to you…

I have yet to see a website that could not be further optimized. It just comes down to what is cheaper, more hardware or more time spent optimizing.


#17

Okay,

Basically, if I send the slide all the way left it works… It’s just that not all the time…

John offer me an option where the sleeping processes are limited to 10… I accept that offer. It should work… John and you has worked well looking into the issue and I appreciate that.

I think the real issue is that my site is very efficient but get visited a lot. Apache then over allocate sleeping processes…

Another issue is that during peak, sometimes I see some of my process has status H. John said, after research, that the CPU is maxed out and hence the status is H… However, the top command also shows that I am using only 10-17% CPU power. So yes CPU is maxed out, but not by me. Looks like someone else using that PS causing all this problem.

Also can I know what’s Dreamhost policy for forcing people to use PS… I mean, CPU is an issue… But do you usually force people to use PS for using too much memory?

I just feel that somehow I am treated differently. Had I not switched to ps, then considering my CPU usage is pretty low, I wouldn’t have ran into this problem.

This thing cut my profit margin significantly but the business is still profitable. Maybe I should buy 2-3 reseller accounts because they are far cheaper than 1 ps… However, I think would like to try regular account for 2 days first before making that decision. Somehow I still believe that my sites are efficient enough it should fit into 1.

This issue has forced me to postpone my expansion plan for long. So if possible, I’d really love to go back and try regular reseller account for just 2 days quickly.

Another that I noticed is that if I slide the cursor to the right, CPU graph drops… I am not sure whether somehow I ended up using less CPU… Looks like another issue is I have a “bad” neighbor which forced me to buy far more than what I use on average to face worst case scenario.

The graph for memory is clear… It has a unit… I know, from the graph that I am currently using between 264 to 691 MB memory… However, from the graph all I can see is I am using 2, 5, or 10… 10 what? 10 MHz? 10% ? Also if I increase the slide, CPU usage actually drops. Now I am using like 2 in the graph. I don’t know. 2 what?

Also the graph for CPU vary wildly… On the other hand top always give a stable number. According to top, CPU usage is always between 10-15% most of the time (the highest I see is 17%)…

Thanks…


#18

I am back to regular server.

Thanks a lot man…

Now how do I verify that I am not causing trouble for you…

Oh I think I know how… Enable tracking and run top…

(Those graphs are kind of cute)…

It’s time to expand… Soon, I’ll probably by a second account… Hopefully not very soon… I have so much unused bandwidth and spaces…

Thanks a lot guys… Your panel is way better than cpanel by the way… That’s why I keep on nagging in this forum… Where else would I go if not to you again?


Hmm… that’s strange, I still have a lot of memory and there are still a lot of dhapache sleeping process… I do believe those dhapache sleeping processes are redundant. That’s why they’re sleeping…

Well, anyway… thanks for allowing me to go back to regular server… My sites are even more stable now by the way…


#19

Hopefully, you won’t need to buy a second account. I haven’t heard of anybody filling up their disk space and using their allocated bandwidth. Anybody who manages to pull that off should probably be on a private server. Just keep adding users and domains to your current account.

-Scott


#20

Yes…

But second account is cheaper than PS… Way cheaper…