Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


BuyVM OpenVZ VPS Review: 512MB/1GB - Centmin Nginx installed
New on LowEndTalk? Please Register and read our Community Rules.

BuyVM OpenVZ VPS Review: 512MB/1GB - Centmin Nginx installed

eva2000eva2000 Member
edited September 2011 in Reviews

Hi guys, my first VPS review on this forum and first time OpenVZ VPS user as well.

About 12 days ago I nabbed 2x BuyVM OpenVZ VPS solely for live testing of my modified Centmin install script which is currently undergoing beta testing and revamp to support Menu mode installation http://vbtechsupport.com/920/. Both BuyVM OpenVZ are 512MB/1GB Burstable/50GB based VPS running CentOS 6.0 32bit.

VPS are on Intel Xeon E5520 Nehalem based processors with 2 cpu cores.

processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 26
model name      : Intel(R) Xeon(R) CPU           L5520  @ 2.27GHz
stepping        : 5
cpu MHz         : 2266.978
cache size      : 4096 KB
physical id     : 0
siblings        : 6
core id         : 0
cpu cores       : 6
apicid          : 0
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc pni ssse3 cx16 sse4_1 sse4_2 popcnt lahf_lm
bogomips        : 4533.95
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 26
model name      : Intel(R) Xeon(R) CPU           L5520  @ 2.27GHz
stepping        : 5
cpu MHz         : 2266.978
cache size      : 4096 KB
physical id     : 0
siblings        : 6
core id         : 1
cpu cores       : 6
apicid          : 1
fpu             : yes
fpu_exception   : yes
cpuid level     : 11
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc pni ssse3 cx16 sse4_1 sse4_2 popcnt lahf_lm
bogomips        : 4532.15
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

Out of the box virgin memory/disk usage and top stats:

             total       used       free     shared    buffers     cached
Mem:          1024         51        972          0          0          0
-/+ buffers/cache:         51        972
Swap:            0          0          0

Filesystem            Size  Used Avail Use% Mounted on
/dev/simfs             50G  566M   50G   2% /
none                  512M  4.0K  512M   1% /dev

top - 14:22:11 up  1:05,  1 user,  load average: 0.00, 0.00, 0.00
Tasks:  16 total,   1 running,  15 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.1%us,  0.2%sy,  0.0%ni, 99.7%id,  0.1%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   1048576k total,    53324k used,   995252k free,        0k buffers
Swap:        0k total,        0k used,        0k free,        0k cached

Disk tests

dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync && rm -rf test
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 22.4788 s, 47.8 MB/s
dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync && rm -rf test
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 7.07835 s, 152 MB/s

My ioping.sh Disk Latency test which tests random disk speed rather than sequential disk speed http://vbtechsupport.com/1239/

-----------------------------------------
ioping.sh 0.9.8 - http://vbtechsupport.com
by George Liu (eva2000)
-----------------------------------------
          ioping.sh 0.9.8 MENU
-----------------------------------------
1. Install ioping
2. Re-install ioping
3. Run ioping default tests
4. Run ioping custom tests
5. Exit
-----------------------------------------
Enter option [ 1 - 5 ] 3
-----------------------------------------
Virtuzzo OR OpenVZ Virtualisation detected

***************************************************
ioping code.google.com/p/ioping/
ioping.sh 0.9.8
shell wrapper script by George Liu (eva2000)
http://vbtechsupport.com
***************************************************

Virtuzzo or OpenVZ Virtualisation detected
**********************************
dd (sequential disk speed test)...
**********************************
dd if=/dev/zero of=testfilex bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 8.87793 s, 121 MB/s

************************
starting ioping tests...
***************************************************
ioping disk I/O test (default 1MB working set)
***************************************************
disk I/O: /
--- / (simfs /dev/simfs) ioping statistics ---
5 requests completed in 4032.8 ms, 2648 iops, 10.3 mb/s
min/avg/max/mdev = 0.3/0.4/0.4/0.0 ms

**********************************************
seek rate test (default 1MB working set)
**********************************************
seek rate: /
--- / (simfs /dev/simfs) ioping statistics ---
370 requests completed in 3004.5 ms, 1475 iops, 5.8 mb/s
min/avg/max/mdev = 0.3/0.7/16.4/1.2 ms

**********************************************
sequential test (default 1MB working set)
**********************************************
-----------------------
sequential: /
--- / (simfs /dev/simfs) ioping statistics ---
104 requests completed in 3018.0 ms, 45 iops, 11.2 mb/s
min/avg/max/mdev = 15.3/22.2/183.5/17.3 ms
-----------------------
sequential cached I/O: /
--- / (simfs /dev/simfs) ioping statistics ---
376 requests completed in 3005.6 ms, 4557 iops, 1139.3 mb/s
min/avg/max/mdev = 0.1/0.2/13.4/0.8 ms

Network speed

wget -O /dev/null http://cachefly.cachefly.net/100mb.test && rm -rf 100mb.test
--2011-09-13 22:33:41--  http://cachefly.cachefly.net/100mb.test
Resolving cachefly.cachefly.net... 205.234.175.175
Connecting to cachefly.cachefly.net|205.234.175.175|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: “/dev/null”

100%[=============================================================================>] 104,857,600 25.0M/s   in 4.2s    

2011-09-13 22:33:46 (24.0 MB/s) - “/dev/null” saved [104857600/104857600]

2nd run

wget -O /dev/null http://cachefly.cachefly.net/100mb.test
--2011-09-14 02:21:31--  http://cachefly.cachefly.net/100mb.test
Resolving cachefly.cachefly.net... 205.234.175.175
Connecting to cachefly.cachefly.net|205.234.175.175|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: “/dev/null”

100%[=============================================================================>] 104,857,600 31.5M/s   in 3.6s    

2011-09-14 02:21:35 (28.1 MB/s) - “/dev/null” saved [104857600/104857600]

Apache v2.2.15 is running by default without any MySQL server installed:

 httpd -V
Server version: Apache/2.2.15 (Unix)
Server built:   Jul  7 2011 11:27:40
Server's Module Magic Number: 20051115:24
Server loaded:  APR 1.3.9, APR-Util 1.3.9
Compiled using: APR 1.3.9, APR-Util 1.3.9
Architecture:   32-bit
Server MPM:     Prefork
  threaded:     no
    forked:     yes (variable process count)
Server compiled with....
 -D APACHE_MPM_DIR="server/mpm/prefork"
 -D APR_HAS_SENDFILE
 -D APR_HAS_MMAP
 -D APR_HAVE_IPV6 (IPv4-mapped addresses enabled)
 -D APR_USE_SYSVSEM_SERIALIZE
 -D APR_USE_PTHREAD_SERIALIZE
 -D SINGLE_LISTEN_UNSERIALIZED_ACCEPT
 -D APR_HAS_OTHER_CHILD
 -D AP_HAVE_RELIABLE_PIPED_LOGS
 -D DYNAMIC_MODULE_LIMIT=128
 -D HTTPD_ROOT="/etc/httpd"
 -D SUEXEC_BIN="/usr/sbin/suexec"
 -D DEFAULT_PIDLOG="run/httpd.pid"
 -D DEFAULT_SCOREBOARD="logs/apache_runtime_status"
 -D DEFAULT_LOCKFILE="logs/accept.lock"
 -D DEFAULT_ERRORLOG="logs/error_log"
 -D AP_TYPES_CONFIG_FILE="conf/mime.types"
 -D SERVER_CONFIG_FILE="conf/httpd.conf"

Comments

  • eva2000eva2000 Member
    edited September 2011

    Before I could run my custom shell scripts, I had to at least install a few items that were missing on virgin VPS.

    yum -y install gcc gcc-c++ file bc perl-TermReadKey perl-Time-HiRes perl-DBD-MySQL nano elinks
    

    Removing Apache v2.2.15 and install Centmin v1.2.1-eva2000.13

    I ran modded Centmin install script which by default installed the following items:

    • Nginx v1.0.6
    • PHP 5.3.8 php-fpm custom source install
    • MariaDB 5.2.7 default MySQL server - uses more memory but faster than MySQL community versions. Benchmarks
    • Memcached 1.4.7 server - 2x 16MB instances including password protected memcache.php admin end
    • Memcache v3.0.5 php extension
    • Xcache v1.3.2 or APC 3.1.9 with respective admin browser interfaces - I choose APC 3.1.9
    • Python v2.72
    • CSF firewall
    • Siege benchmark and Sproxy

    Centmin v1.2.1-eva2000.13 beta
    Centmin menu version's options

    --------------------------------------------------------
    Centmin 1.2.1-eva2000.13 - http://vbtechsupport.com/920/
    Menu/Mods Author: eva2000 (vbtechsupport.com)
    Centmin Original Author: BTCentral (btcentral.org.uk)
    --------------------------------------------------------
                       Centmin Menu                   
    --------------------------------------------------------
    1).  Centmin Install
    2).  Add Nginx vhost domain
    3).  NSD setup domain name DNS
    4).  Nginx Upgrade
    5).  PHP Upgrade
    6).  XCache Re-install
    7).  APC Cache Re-install
    8).  XCache Install
    9).  APC Cache Install
    10). Memcached Server Re-install
    11). MariaDB 5.2 Upgrade
    12). Install ioping.sh vbtechsupport.com/1239/
    13). SELinux disable
    14). Exit
    --------------------------------------------------------
    Enter option [ 1 - 14 ] 
    --------------------------------------------------------
    

    I usually test my modded Centmin script on local virtualbox based CentOS 5.5/5.6/6.0 64bit servers. This was first time testing on OpenVZ VPS server and was surprised to see alot higher memory usage right after installation, particularly for Memcached servers - 2x 16MB instances took nearly 54MB each. Reading up I found it's due to OpenVZ's memory usage and default system stack sizes.

    From http://code.google.com/p/memcached/issues/detail?id=186

    On OpenVZ (operating-system level virtualization for Linux), there is no differentiation between virtual and resident memory, so all the virtual memory is counted as RAM. The default installation of Memcached therefore uses up 50MB of RAM (4+1 threads x 10MB) when starting up on an OpenVZ virtual private server.

    I can confirm that this was the case on my new test buyvm OpenVZ 512MB/1GB burstable CentOS 6.0 32bit VPS. This CentOS 6.0 32bit based OpenVZ VPS has a default system stack size at 10240KB.

    I wrote a quick servicemem.sh bash shell script to gather all info and tests for the below output where the script outputs base memory usage, then it stops/starts each key service and takes a free -m output after each key service has stopped/started.

    As you can see on this fresh virgin Centmin v1.2.1-eva2000.13 beta Nginx/PHP-FPM installed, this OpenVZ VPS shows the following amount of memory used by each of the listed services

    1. with default system stack size at 10240KB
    2. with tweaked service init.d stack sizes
    3. with tweaked service init.d stack size + lighter MariaDB mysql my.cnf and apc.ini settings. Note MariaDB has higher mem usage than MySQL community versions as it has a new Aria storage engine which is needed to be running as internal on-disk tables are now in Aria format instead of MyISAM which helps speed up MySQL query performance where GROUP BY and DISTINCT occur. Read http://kb.askmonty.org/en/aria-storage-engine. I lowered the Aria log file size as well as Aria Pagecache and Sort buffer sizes from defaults which may lead to less performance.

    1) Default system stack size

    base mem usage
                 total       used       free     shared    buffers     cached
    Mem:          1024        328        695          0          0          0
    -/+ buffers/cache:        328        695
    Swap:            0          0          0
    
    • rsyslod = 29MB
    • lfd = 10MB
    • php-fpm = 44MB
    • nginx = 30MB
    • mysql = 91MB
    • memcached = 108MB

    2) After tweaked system stack size for specific service init.d scripts

    base mem usage
                 total       used       free     shared    buffers     cached
    Mem:          1024        200        823          0          0          0
    -/+ buffers/cache:        200        823
    Swap:            0          0          0
    
    • rsyslod = 3MB
    • lfd = 9MB
    • php-fpm = 42MB
    • nginx = 30MB
    • mysql = 88MB
    • memcached = 6MB

    using the following stack sizes

    ##################################
    # nano -w /etc/init.d/memcached
    # ulimit -s 256
    # service memcached restart
    ##################################
    # nano -w /etc/init.d/nginx
    # ulimit -s 128
    # service nginx restart
    ##################################
    # 256 for 64bit & 192 for 32bit
    # nano -w /etc/init.d/mysql
    # service mysql restart
    ##################################
    # nano -w /etc/init.d/php-fpm
    # ulimit -s 512
    # service php-fpm restart
    ##################################
    # nano -w /etc/init.d/rsyslog
    # ulimit -s 512
    # service rsyslog restart
    ##################################
    # nano -w /etc/init.d/lfd
    # ulimit -s 2048
    # service lfd restart
    ##################################
    

    3) Tuned MariaDB my.cnf and APC php.ini memory allocation

    base mem usage
                 total       used       free     shared    buffers     cached
    Mem:          1024        114        909          0          0          0
    -/+ buffers/cache:        114        909
    Swap:            0          0          0
    
    • rsyslod = 3MB
    • lfd = 9MB
    • php-fpm = 24MB
    • nginx = 14MB
    • mysql = 39MB
    • memcached = 5MB
  • Uhm, very nice and details review. BuyVM is a good budget vps provider out there =D

  • eva2000eva2000 Member
    edited September 2011

    Thanks giang..

    Geekbench v2.2.0 = 3,655
    http://browse.geekbench.ca/geekbench2/view/481627

    vs

    Geekbench v2.2.1 beta = 3,045
    http://browse.geekbench.ca/geekbench2/view/481628

    Here's a static html file siegebench.sh test run from local BuyVM VPS vs remote Wiredtree VPS to target server (BuyVM VPS).

    local test = running siegebench.sh on same server as target url's test BuyVM VPS server

    • buyvm OpenVZ 512MB/1GB Burstable VPS (2 core Intel Xeon E5520)

    vs

    remote = running siegebench.sh on remote server testing target url's test BuyVM VPS server

    • wiredtree Virtouzzo 1.5GB WHM/Cpanel VPS (shared 8 core Dual Intel Xeon E5405)

    local

    Thu Sep 15 12:19:40 MSD 2011
    Log files saved at /home/siegebenchlogs
                           
    Sep 15  12:19   947    siegehtml_benchmark_150911-1218.txt
    
    -------------------------------------------------
    contents of /home/siegebenchlogs/siegehtml_benchmark_150911-1218.txt
    -------------------------------------------------
    date | trans | elapse time | data trans (MB) | response time | trans/sec | throughput (MB/s) | concurrency | success trans | failed trans
    -------------------------------------------------
    **** siege_concurrentusers_50 ****
    2011-09-15 12:18:22,   2500,       1.72,           0,       0.02,     1453.49,        0.00,       32.32,    2500,       0
    **** siege_concurrentusers_100 ****
    2011-09-15 12:18:32,   5000,       3.26,           0,       0.04,     1533.74,        0.00,       62.81,    5000,       0
    **** siege_concurrentusers_150 ****
    2011-09-15 12:18:43,   7500,       5.42,           0,       0.08,     1383.76,        0.00,      107.63,    7500,       0
    **** siege_concurrentusers_200 ****
    2011-09-15 12:18:56,  10000,       6.62,           1,       0.08,     1510.57,        0.15,      125.97,   10000,       0
    **** siege_concurrentusers_250 ****
    2011-09-15 12:19:13,  12500,      11.20,           1,       0.15,     1116.07,        0.09,      172.27,   12500,       0
    **** siege_concurrentusers_300 ****
    2011-09-15 12:19:34,  15000,      14.71,           1,       0.17,     1019.71,        0.07,      178.32,   15000,       0
    -------------------------------------------------
    

    remote

    Thu Sep 15 03:25:46 CDT 2011
    Log files saved at /home/siegebenchlogs
                           
    Sep 15  03:25   947    siegehtml_benchmark_150911-0324.txt
    
    -------------------------------------------------
    contents of /home/siegebenchlogs/siegehtml_benchmark_150911-0324.txt
    -------------------------------------------------
    date | trans | elapse time | data trans (MB) | response time | trans/sec | throughput (MB/s) | concurrency | success trans | failed trans
    -------------------------------------------------
    **** siege_concurrentusers_50 ****
    2011-09-15 03:24:14,   2500,       5.86,           0,       0.12,      426.62,        0.00,       49.08,    2500,       0
    **** siege_concurrentusers_100 ****
    2011-09-15 03:24:29,   5000,       8.77,           0,       0.12,      570.13,        0.00,       65.68,    5000,       0
    **** siege_concurrentusers_150 ****
    2011-09-15 03:24:45,   7500,       9.19,           0,       0.12,      816.10,        0.00,       95.34,    7500,       0
    **** siege_concurrentusers_200 ****
    2011-09-15 03:24:57,  10000,       6.75,           1,       0.12,     1481.48,        0.15,      175.56,   10000,       0
    **** siege_concurrentusers_250 ****
    2011-09-15 03:25:18,  12500,      14.90,           1,       0.20,      838.93,        0.07,      169.21,   12500,       0
    **** siege_concurrentusers_300 ****
    2011-09-15 03:25:40,  15000,      15.95,           1,       0.22,      940.44,        0.06,      207.30,   15000,       0
    -------------------------------------------------
    
  • Lengthy, but accurate review. Hmmmm... but is this actually a review of BuyVM or what you did to get your service to work? :p

  • This is definitely the most detailed review of a VPS I've ever seen! And Buyvm seems to be doing well through all these tests. GJ Fran!

  • thanks fan :)

    Boltersdriveer said: Lengthy, but accurate review. Hmmmm... but is this actually a review of BuyVM or what you did to get your service to work? :p

    why can't it be about both ? :D

    Geekbench v2.2.0 = 3,655
    http://browse.geekbench.ca/geekbench2/view/481627

    vs

    Geekbench v2.2.1 beta = 3,045
    http://browse.geekbench.ca/geekbench2/view/481628

  • yomeroyomero Member
    edited September 2011

    Boltersdriveer said: or what you did to get your service to work?

    Seems more like that

  • Grabbed my 3rd buyvm VPS recently to build my multi-server testing cluster which will be load balanced + tungsten replicator mysql replication based. Just waiting on buyvm to release more VPS on separate nodes from the VPS i currently have :)

  • I run similar stuff on my web box eva2000 (who here doesn't, probably). Have you tried Percona MySQL, and do you feel MariaDB is better than it if so?

    Also, you should try out the xCache beta 2, I always liked eAccelerator the best, then that died, and then APC became my favorite, but this xCache beta performs so much better. It's stable as well, and seems to use quite a bit less memory.

  • Kairus, yup tried MySQL community releases, Percona 5.1/5.5 and MariaDB 5.2 and MariaDB 5.2 is the best of them all http://vbtechsupport.com/606/ and http://vbtechsupport.com/657/. Why because MariaDB 5.2 uses Percona's XtraDB InnoDB engine so has same or somewhat better InnoDB performance as Percona but in addition MariaDB is the only MySQL version which still focuses on MySQL core improvements as well as improvements to MyISAM engine.

    Most of all the clients and sites I have dealt with are more MyISAM based i.e. wordpress and vBulletin. Also InnoDB memory and storage requirements are much higher than MyISAM for optimal performance, so even for LEB with small memory and disk space allocations, InnoDB wouldn't be the ideal storage engine to choose.

    MariaDB 5.3 is in beta but is the next level of performance back port MySQL 5.5/5.6 features and making further performance improvements especially in MySQL replication.

    As to Xcache haven't tried beta 2.x yet, but definitely will try it out - thanks for the heads up :)

    Put my 3x buyvm OpenVZ VPS to work at http://centminmod.com next LEB VPS I get will have to be KVM/Xen based as I need to be able to make sysctl.conf changes on load balancer end for failover :)

  • fanfan Member

    Any plan for a script for Debian based OS's?

  • fan said: Any plan for a script for Debian based OS's?

    Been playing with Debian too mainly just to make sure MariaDB 5.2 installs and optimised properly for Debian based clients of mine but yes I can see Debian version of centmin in the future. Just not right now, centmin mod version still has alot of stuff outstanding I want to add to it - including phpmyadmin behind SSL and separate uid/guid php-fpm pools for phpmyadmin vhost as well as adding Percona 5.5 MySQL and MySQL 5.5 support to centmin :)

    Centmin for php and nginx upgrade functions has error checking at configure, make and make install stages of upgrades so easier to check if any stages fail and why. But haven't added the checks to initial install functions yet so that's another to do list item.

    Might add mysqlmymonlite.sh and mysqlmybackup.sh support to centmin too later on.

  • fanfan Member

    eva2000 said: Been playing with Debian too mainly just to make sure MariaDB 5.2 installs and optimised properly for Debian based clients of mine but yes I can see Debian version of centmin in the future. Just not right now, centmin mod version still has alot of stuff outstanding I want to add to it - including phpmyadmin behind SSL and separate uid/guid php-fpm pools for phpmyadmin vhost as well as adding Percona 5.5 MySQL and MySQL 5.5 support to centmin :)

    Anyway, excellent script! Maybe I'll reload one of my servers to Debian to check it out. :P

  • Thanks. You mean reload with Centos ? :)

Sign In or Register to comment.