speed of a php code

[alexks]

I have php code for inserting data from csv file to database. It runs for 10 min on PC and for 30 min on RHEL.

Does anybody know what causes this delay on RHEL? Can this be php's configuration problem?

Thanks.
A.

[nrg_alpha]

I think the first thing people are going to want to know is what the code looks like.
This would be helpful.. Right now, it's sort of like telling the shop mechanic that your car's engine is under performing and making strange sounds, but not allowing the him/her to pop open the hood to take a look.

Cheers,

NRG

[bretticus]

Even 10 minutes is a long time. Must be a huge file! Perhaps your PC has more RAM and doesn't have to swap memory as often as your RHEL server.

[alexks]

The csv file has about 20,000 lines. I increased memory in RHEL 8 times but there is no change in speed at all. So, I thought maybe php configuration is different in RHEL, but I am not sure which configuration can cause this problem.

[alexks]

nrg_alpha wrote:

I think the first thing people are going to want to know is what the code looks like.

The code is a simple sql and php to get data line by line from csv file and insert it to database. It works without any errors in PC and RHEL. But as I said it is slower in RHEL, and this is not memory problem.

[illzz]

How do the processors compare?

[illzz]

Are both web servers connecting to the same database server? Is it remote to both of them?.

What does top look like when it is processing? Who's eating up the resource(s)? And which resource(s)?

[alexks]

The databases are local for both PC and RHEL. In RHEL I have vm and a database. Everything happens inside the vm.

illzz wrote:

What does top look like when it is processing? Who's eating up the resource(s)? And which resource(s)?

Not sure what all of these mean? How can I find who is eating resources?

Thanks.
A.

[Kudose]

At your linux shell console type

top

Hit enter

[illzz]

top is a *NIX command-line tool to see resource usage in real time (along with a bunch of other things)

look at the man page for more information, such as sorting.

basically you can sort by CPU% by pressing P (capital), or MEM% by pressing M (capital). You can also force a refresh (defaults to 3 seconds, i think) by hitting the space bar.

Try looking at that while the PHP script is running to see if it is actually PHP using all the resources, or if it something else, such as MySQL. Then you can see where the bottle neck is (CPU or MEM).

[illzz]

Also... is your Windows running inside of a VM as well?

I've encountered some performance hits running inside of vmware.

[alexks]

Windows is running on a different computer. Here is the result of top

top - 14:07:15 up 4:01, 3 users, load average: 1.15, 0.31, 0.10
Tasks: 99 total, 3 running, 96 sleeping, 0 stopped, 0 zombie
Cpu(s): 67.3% us, 32.7% sy, 0.0% ni, 0.0% id, 0.0% wa, 0.0% hi, 0.0% si
Mem: 1554728k total, 281060k used, 1273668k free, 32000k buffers
Swap: 511992k total, 0k used, 511992k free, 131020k cached

PID USER PR NI %CPU TIME+ %MEM VIRT RES SHR S COMMAND

3343 mysql 16 0 81.0 33:34.75 1.8 169m 27m 3416 S mysqld

9818 som 25 0 18.6 0:13.98 2.3 186m 35m 13m R php

2345 root 16 0 0.3 0:30.98 0.0 1872 492 396 R vmware-guestd

9820 som 16 0 0.3 0:00.21 0.1 2040 1008 780 R top

1 root 16 0 0.0 0:00.84 0.0 3340 548 468 S init

2 root RT 0 0.0 0:00.00 0.0 0 0 0 S migration/0

3 root 34 19 0.0 0:00.00 0.0 0 0 0 S ksoftirqd/0

4 root 5 -10 0.0 0:00.03 0.0 0 0 0 S events/0

5 root 5 -10 0.0 0:00.01 0.0 0 0 0 S khelper

6 root 15 -10 0.0 0:00.00 0.0 0 0 0 S kacpid

17 root 5 -10 0.0 0:00.02 0.0 0 0 0 S kblockd/0

18 root 25 0 0.0 0:00.00 0.0 0 0 0 S khubd

35 root 20 0 0.0 0:00.00 0.0 0 0 0 S pdflush

36 root 15 0 0.0 0:01.03 0.0 0 0 0 S pdflush

37 root 25 0 0.0 0:00.00 0.0 0 0 0 S kswapd0

38 root 11 -10 0.0 0:00.00 0.0 0 0 0 S aio/0

184 root 25 0 0.0 0:00.00 0.0 0 0 0 S kseriod

416 root 23 0 0.0 0:00.00 0.0 0 0 0 S scsi_eh_0

443 root 16 0 0.0 0:00.47 0.0 0 0 0 S kjournald

1087 root 6 -10 0.0 0:00.00 0.0 0 0 0 S kauditd

1638 root 7 -10 0.0 0:00.00 0.0 0 0 0 S ata/0

1639 root 7 -10 0.0 0:00.00 0.0 0 0 0 S ata_aux

1657 root 6 -10 0.0 0:00.07 0.0 1664 440 368 S udevd

1813 root 23 0 0.0 0:00.00 0.0 0 0 0 S kjournald

1814 root 16 0 0.0 0:00.45 0.0 0 0 0 S kjournald

2202 root 16 0 0.0 0:00.26 0.0 2888 528 440 S syslogd

2206 root 16 0 0.0 0:00.02 0.0 2500 376 312 S klogd

[illzz]

So it looks like your CPU is being pegged by MySQL at that snapshot. (81% by MySQL, 18.6% by PHP). Maybe keep an eye on it for a while to make sure that pattern is fairly consistent, but I would guess that is your problem.

You are saying that Windows is running on a different computer, but is it running inside of vmware as well? Or is it running native?

Also, you never said how the processors compared? Are they the same type? speed? etc?

[alexks]

How can I fix mysql's taking 81% of the memeory. Why it does not take so much on Windows? Windows is running native. No vms.

illzz wrote:

Also, you never said how the processors compared? Are they the same type? speed? etc?

For windows, I have Core (TM)2 Duo CPU T7300 #2.00GHz 1.18 GHz 1.99 GB of RAM

For RHEL, I am not sure which processor we have, but I put memory for vm 1536MB
But this increase did not change the speed of the script even for 1 sec.

[illzz]

your memory doesn't look like the problem, it looks like it's the CPU. From the snapshot, MySQL was only using 1.8% of the available "physical" memory. (what vmware is presenting as physical)

As for the CPU stats, run cat /proc/cpuinfo and post the output.

The problem could be that you are running RHEL in a VM whereas the Windows version is running native. I bet that if you are run RHEL native, you would get better performance.

[alexks]

cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 15
model : 4
model name : Intel(R) Pentium(R) 4 CPU 3.20GHz
stepping : 8
cpu MHz : 3188.817
cache size : 2048 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 5
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss constant_tsc pni ds_cpl
bogomips : 4718.59

[illzz]

You may want to also compare the configurations for the 2 databases to make sure there aren't any major differences.

Would you be able to test performance with RHEL running the web server, but connecting to the Windows MySQL DB?

Are there any other VMs running on the same machine as the one hosting the RHEL?

Do you have any instance of RHEL that is not in a VM, that is available for testing performance?

[alexks]

Not sure what do you mean by configuration of databases? Both of them are mysql, the same tables. On RHEL there is a second vm, but it is stopped. I have native RHEL, but in order to test this code, I must install all these php apache and mysql. I kept an eye on top results, Mysql was using even 96 % of CPU. The lowest was 86%.

[illzz]

the configuration files are typically named either my.cnf or my.ini... there may be some optimization settings in one server but not the other...

What about running the web server on RHEL connecting to the database on Windows? I have a feeling that vmware may be part of the problem you are having. There is a performance hit on running an OS on virtualized hardware vs running it on native hardware. You can change some settings to minimize this, but comparable native hardware typically runs better than it's virtualized brethren.