Administrators · Legacy archive
Site Upgrade
-
Site Upgrade
#1I put in the support request to boost us to our own private server.
Lets see what happens........ -
Re: Site Upgrade
#2*AG prepares to chase new, weird encoding from his posts* -
Re: Site Upgrade
#3Should be seemless as it is the same host vs. me having to gzip and resetup. -
Re: Site Upgrade
#4Stuff for me and you to compare. We got a pretty good one, unloaded.
http://www.webhostingtalk.com/showthread.php?t=409584
Start Benchmark Run: Sat Apr 15 10:57:49 CDT 2006
10:57:49 up 3:40, 1 user, load average: 0.02, 0.01, 0.00
End Benchmark Run: Sat Apr 15 11:10:02 CDT 2006
11:10:02 up 3:53, 1 user, load average: 11.91, 5.43, 2.59
INDEX VALUES
TEST BASELINE RESULT INDEX
Dhrystone 2 using register variables 376783.7 3069280.9 81.5
Double-Precision Whetstone 83.1 944.9 113.7
Execl Throughput 188.3 1024.0 54.4
File Copy 1024 bufsize 2000 maxblocks 2672.0 14850.0 55.6
File Copy 256 bufsize 500 maxblocks 1077.0 4381.0 40.7
File Read 4096 bufsize 8000 maxblocks 15382.0 117978.0 76.7
Pipe Throughput 111814.6 130842.5 11.7
Pipe-based Context Switching 15448.6 44640.6 28.9
Process Creation 569.3 4371.2 76.8
Shell Scripts (8 concurrent) 44.8 237.9 53.1
System Call Overhead 114433.5 169058.1 14.8
=========
FINAL SCORE 45.6
-
Re: Site Upgrade
#5I don't know what some of this means but doing a before and after in case I ever figure it out.
cat /proc/user_beancounters
Version: 2.5
uid resource held maxheld barrier limit failcnt
4220: kmemsize 4024883 4042380 54156000 54156000 0
lockedpages 0 0 4096 4096 0
privvmpages 47528 47537 14680064 14680064 0
shmpages 859 859 131072 131072 0
dummy 0 0 0 0 0
numproc 61 61 400 400 0
physpages 32548 32549 0 2147483647 0
vmguarpages 0 0 14680064 2147483647 0
oomguarpages 32549 32550 107520 2147483647 0
numtcpsock 26 26 500 500 0
numflock 6 6 200 220 0
numpty 1 1 64 64 0
numsiginfo 0 0 512 512 0
tcpsndbuf 4456 11140 5365760 10485760 0
tcprcvbuf 0 692 5365760 10485760 0
othersockbuf 28528 28528 1503232 4063232 0
dgramrcvbuf 0 0 262144 262144 0
numothersock 23 23 500 500 0
dcachesize 449010 451880 12288000 12288000 0
numfile 1614 1615 32000 32000 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
numiptent 18 18 200 200 0
VMGUARPAGES
This parameter controls how much memory is available to VPS (i.e.how much memory its applications can allocate by 'malloc(3)' or other standard Linux memory allocation mechanisms). The more clients are served or the more ``heavy'' the application is, the more memory it needs. The amount of memory that Virtual Private Server's applications are guaranteed to be able to allocate is specified as the 'barrier' of 'vmguarpages' parameter. The current amount of allocated memory space is accounted into 'privvmpages' parameter, and 'vmguarpages' parameter doesn not have its own accounting.
PRIVVMPAGES
'Privvmpages' allows controlling the amount of memory allocated by applications. The 'barrier' and the 'limit' of parameter control the upper boundary of the total size of allocated memory. Note that this upper boundary doesn't guarantee that the Virtual Private Server will be able to allocate that much memory, neither does it guarantee that other Virtual Private Servers will be able to allocate their fair share of memory. The primary mechanism to control memory allocation is the 'vmguarpages' guarantee. 'Privvmpages' parameter accounts allocated (but, possibly, not used yet) memory. It is important not to allow any Virtual Private Server to allocate a significant portion of all system RAM to avoid serious service level degradation for other environments.
SHMPAGES
These pages are also accounted into 'privvmpages'. The configuration of this parameter doesn't affect security and stability of the whole system or isolation between Virtual Private Servers. Its configuration affects functionality and resource shortage reaction of applications in the given Virtual Private Server only.
NUMPROC
Many server applications (like Apache Web server, FTP and mail servers) spawn a process to handle each client, so the limit on number of processes defines how many clients the application will be able to handle in parallel. However, the number of processes doesn't limit how ``heavy'' the application is and whether the server will be able to serve heavy requests from clients. -
Re: Site Upgrade
#6free
total used free shared buffers cached
Mem: 12443296 12353048 90248 0 107060 3975340
-/+ buffers/cache: 8270648 4172648
Swap: 16771728 1396688 15375040
-
Re: Site Upgrade
#7Current Box
Start Benchmark Run: Thu Aug 17 19:49:56 CDT 2006
19:49:56 up 7 days, 12:07, 1 user, load average: 0.19, 0.20, 0.12
End Benchmark Run: Thu Aug 17 20:02:17 CDT 2006
20:02:17 up 7 days, 12:19, 1 user, load average: 15.00, 6.55, 3.10
INDEX VALUES
TEST BASELINE RESULT INDEX
Dhrystone 2 using register variables 376783.7 3206203.6 85.1
Double-Precision Whetstone 83.1 1031.5 124.1
Execl Throughput 188.3 1535.3 81.5
File Copy 1024 bufsize 2000 maxblocks 2672.0 19716.0 73.8
File Copy 256 bufsize 500 maxblocks 1077.0 9289.0 86.2
File Read 4096 bufsize 8000 maxblocks 15382.0 190712.0 124.0
Pipe Throughput 111814.6 225024.8 20.1
Pipe-based Context Switching 15448.6 69222.6 44.8
Process Creation 569.3 7566.6 132.9
Shell Scripts (8 concurrent) 44.8 344.9 77.0
System Call Overhead 114433.5 195850.1 17.1
=========
FINAL SCORE 66.2
------------------------
Version: 2.5
uid resource held maxheld barrier limit failcnt
4220: kmemsize 6355692 6373189 54156000 54156000 0
lockedpages 0 0 4096 4096 0
privvmpages 74124 74132 14680064 14680064 0
shmpages 4827 4827 131072 131072 0
dummy 0 0 0 0 0
numproc 77 77 400 400 0
physpages 55332 55333 0 2147483647 0
vmguarpages 0 0 14680064 2147483647 0
oomguarpages 61642 61643 107520 2147483647 0
numtcpsock 31 31 500 500 0
numflock 3 3 200 220 0
numpty 1 1 64 64 0
numsiginfo 2 3 512 512 0
tcpsndbuf 4456 6684 5365760 10485760 0
tcprcvbuf 0 4276 5365760 10485760 0
othersockbuf 26300 29220 1503232 4063232 0
dgramrcvbuf 0 0 262144 262144 0
numothersock 24 26 500 500 0
dcachesize 639436 643454 12288000 12288000 0
numfile 1643 1644 32000 32000 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
numiptent 73 73 200 200 0
--------------------
total used free shared buffers cached
Mem: 16614656 16580240 34416 0 375132 4628132
-/+ buffers/cache: 11576976 5037680
Swap: 16771728 487076 16284652
-
Myriad Networks Test
#8New Server - unloaded
Start Benchmark Run: Fri Aug 18 16:20:59 EDT 2006
16:20:59 up 1:56, 0 users, load average: 0.00, 0.00, 0.00
End Benchmark Run: Fri Aug 18 16:31:24 EDT 2006
16:31:24 up 2:06, 0 users, load average: 14.93, 6.17, 2.68
INDEX VALUES
TEST BASELINE RESULT INDEX
Dhrystone 2 using register variables 376783.7 16437908.8 436.3
Double-Precision Whetstone 83.1 913.7 110.0
Execl Throughput 188.3 3496.2 185.7
File Copy 1024 bufsize 2000 maxblocks 2672.0 37865.0 141.7
File Copy 256 bufsize 500 maxblocks 1077.0 14344.0 133.2
File Read 4096 bufsize 8000 maxblocks 15382.0 286174.0 186.0
Pipe Throughput 111814.6 606753.3 54.3
Pipe-based Context Switching 15448.6 66005.4 42.7
Process Creation 569.3 9468.4 166.3
Shell Scripts (8 concurrent) 44.8 865.7 193.2
System Call Overhead 114433.5 704910.1 61.6
=========
FINAL SCORE 127.4
-------------------------
Version: 2.5
uid resource held maxheld barrier limit failcnt
552: kmemsize 4796749 4814246 23317184 25317184 0
lockedpages 0 0 3686 4096 0
privvmpages 75792 75801 256000 286720 0
shmpages 1371 1371 133120 133120 0
dummy 0 0 0 0 0
numproc 76 76 300 300 0
physpages 29383 29385 0 2147483647 0
vmguarpages 0 0 102400 2147483647 0
oomguarpages 29505 29507 102400 2147483647 0
numtcpsock 26 26 500 500 0
numflock 9 9 200 200 0
numpty 2 2 16 16 0
numsiginfo 0 0 512 512 0
tcpsndbuf 6684 13368 5120000 10240000 0
tcprcvbuf 0 2228 5120000 10240000 0
othersockbuf 32984 32984 2304000 3807232 0
dgramrcvbuf 0 0 1843200 2048000 0
numothersock 27 27 500 500 0
dcachesize 457620 460490 4194304 4317184 0
numfile 1834 1835 8192 8192 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
numiptent 18 18 3000 3000 0
--------------------
total used free shared buffers cached
Mem: 8297404 8035456 261948 0 388580 6156088
-/+ buffers/cache: 1490788 6806616
Swap: 8193140 1204 8191936
-
Re: Site Upgrade
#9Some difference, but you might want to throw in an explanation for the non techs! -
Re: Site Upgrade
#10yes please -
Re: Site Upgrade
#11The numbers that really matter.
Old:
FINAL SCORE 66.2
New:
FINAL SCORE 127.4
(Yes, bigger is better and that is a HUGE jump)
The rest are just empty comparisons really to see where the point spread comes from. Some people know how to read those so I put them there in case I ever care. Before and After and all that.
(And I just got 4 "Cannot Connect" messages in a row trying to preview this. I will move tomorrow or Sunday at the latest! Because they are both servers it is about 10000000 times easier to move them. There will be no extra odd characters that happened with the last moves either because it is done totally different.) -
Re: Site Upgrade
#12Yes. This site has had connection problems more often than I care to mention. Sometimes for as long as 24 hours. For a good server, it sure had its quirks.
When my economy has recovered fully from several years of student salary (which means when all my bills are paid) I will spit something into the donation fund. You deserve it. -
Re: Site Upgrade
#13Ditto :D