polarhome datacenter

[zoli]

polarhome network backbone is upgraded to Mikrotik CCR1009-7G-1C-1S+, TP-LINK TL-SG2424, Cisco Catalyst WS-C2960G-48TC-L. The routers/swithes are second hand, close to end of support date products - but still much better than earlier. The old Mikrotik RB450G, that served well for many years is now in the museum.

I don't brag - but sometimes I feel that it is a pure miracle how polarhome is able to be mostly online while I am working full time - often abroad...hmmm in a server room where yearly temperatures vary between -10 and +50 degrees celsius.

I reveal the secret of the remote management of the small network infrastructure

1. fast, reliable, secure and easily managable router - my choice was Mikrotik
2. stable, remote managable backbone switches that just simply work whatever happens with them. TP-LINK and Cisco have proved to work well in the polarhome's extreme environment
3. Remote managable KVM - Raritan Dominion KX is the heart of the console work. It handles serial consoles, text and graphical ones perfectly.
4. Virtualize whatever you can - easier hadware management, backup, restore etc. - VirtualBox and SIMH was my choice
5. Create periodic backups - modern hard disks live no longer than 3 years with polarhome type of usage. Be prepared.

That's it. This simply works.

Seems the most important that I learned during the past 17 years is: it is better to get what is needed, than to patch and hassle arround with some equipment that it not made for the purpose you intend to use.

[sjaz]

Amazing

[zoli]

Evaluating current and future hypervisors (in polarhome's budget constraints) here are the benchmark results

HP Proliant DL360p G8
CPU: 2 x Intel Xeon Okta Core E5 2690 - 2,9GHz

Code: Select all

Arithmetic Test (type = double)               1.0 lps   (0 secs, 10 samples)
System Call Overhead                     703247.8 lps   (10 secs, 10 samples)
Pipe Throughput                          844713.8 lps   (10 secs, 10 samples)
Pipe-based Context Switching             220941.4 lps   (10 secs, 10 samples)
Process Creation                           5696.6 lps   (30 secs, 3 samples)
Execl Throughput                           no measured results
File Read 1024 bufsize 2000 maxblocks    1635237.0 KBps  (30 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks   418058.0 KBps  (30 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks    319221.0 KBps  (30 secs, 3 samples)
File Read 256 bufsize 500 maxblocks      424542.0 KBps  (30 secs, 3 samples)
File Write 256 bufsize 500 maxblocks     105563.0 KBps  (30 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks       82926.0 KBps  (30 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks    4267863.0 KBps  (30 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks   1346765.0 KBps  (30 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks    915642.0 KBps  (30 secs, 3 samples)
Shell Scripts (1 concurrent)               6748.3 lpm   (60 secs, 3 samples)
Shell Scripts (8 concurrent)               3117.7 lpm   (60 secs, 3 samples)
Shell Scripts (16 concurrent)              1943.3 lpm   (60 secs, 3 samples)
Arithmetic Test (type = short)                1.0 lps   (0 secs, 3 samples)
Arithmetic Test (type = int)                  1.0 lps   (0 secs, 3 samples)
Arithmetic Test (type = long)                 1.0 lps   (0 secs, 3 samples)
Arithmetic Test (type = float)                1.0 lps   (0 secs, 3 samples)
Arithoh                                       1.0 lps   (0 secs, 3 samples)
C Compiler Throughput                      1457.7 lpm   (60 secs, 3 samples)
Dc: sqrt(2) to 99 decimal places         110844.0 lpm   (30 secs, 3 samples)
Recursion Test--Tower of Hanoi           206073.4 lps   (20 secs, 3 samples)


                     INDEX VALUES
TEST                                        BASELINE     RESULT      INDEX

Arithmetic Test (type = double)              29820.0        1.0        0.0
Dhrystone 2 using register variables        116700.0 23334346.0     1999.5
Execl Throughput                                43.0        0.0        0.0
File Copy 1024 bufsize 2000 maxblocks         3960.0   319221.0      806.1
File Copy 256 bufsize 500 maxblocks           1655.0    82926.0      501.1
File Copy 4096 bufsize 8000 maxblocks         5800.0   915642.0     1578.7
Pipe Throughput                              12440.0   844713.8      679.0
Process Creation                               126.0     5696.6      452.1
Shell Scripts (8 concurrent)                     6.0     3117.7     5196.2
System Call Overhead                         15000.0   703247.8      468.8
                                                                 =========
     FINAL SCORE                                                       0.0


Dell PowerEdge R610
CPU: 2 x Intel Xeon Hexa Core X5675 - 3,06GHz

Code: Select all

Arithmetic Test (type = double)               1.0 lps   (0 secs, 10 samples)
System Call Overhead                     840080.5 lps   (10 secs, 10 samples)
Pipe Throughput                          1103127.1 lps   (10 secs, 10 samples)
Pipe-based Context Switching             178540.2 lps   (10 secs, 10 samples)
Process Creation                           3293.4 lps   (30 secs, 3 samples)
Execl Throughput                           no measured results
File Read 1024 bufsize 2000 maxblocks    2078699.0 KBps  (30 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks   1184447.0 KBps  (30 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks    704059.0 KBps  (30 secs, 3 samples)
File Read 256 bufsize 500 maxblocks      558794.0 KBps  (30 secs, 3 samples)
File Write 256 bufsize 500 maxblocks     290331.0 KBps  (30 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks      196860.0 KBps  (30 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks    6992147.0 KBps  (30 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks   4229455.0 KBps  (30 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks    2303576.0 KBps  (30 secs, 3 samples)
Shell Scripts (1 concurrent)               3063.7 lpm   (60 secs, 3 samples)
Shell Scripts (8 concurrent)               1577.0 lpm   (60 secs, 3 samples)
Shell Scripts (16 concurrent)              1109.0 lpm   (60 secs, 3 samples)
Arithmetic Test (type = short)                1.0 lps   (0 secs, 3 samples)
Arithmetic Test (type = int)                  1.0 lps   (0 secs, 3 samples)
Arithmetic Test (type = long)                 1.0 lps   (0 secs, 3 samples)
Arithmetic Test (type = float)                1.0 lps   (0 secs, 3 samples)
Arithoh                                       1.0 lps   (0 secs, 3 samples)
C Compiler Throughput                       979.3 lpm   (60 secs, 3 samples)
Dc: sqrt(2) to 99 decimal places          50906.9 lpm   (30 secs, 3 samples)
Recursion Test--Tower of Hanoi           259061.5 lps   (20 secs, 3 samples)


                     INDEX VALUES
TEST                                        BASELINE     RESULT      INDEX

Arithmetic Test (type = double)              29820.0        1.0        0.0
Dhrystone 2 using register variables        116700.0 27594398.1     2364.6
Execl Throughput                                43.0        0.0        0.0
File Copy 1024 bufsize 2000 maxblocks         3960.0   704059.0     1777.9
File Copy 256 bufsize 500 maxblocks           1655.0   196860.0     1189.5
File Copy 4096 bufsize 8000 maxblocks         5800.0  2303576.0     3971.7
Pipe Throughput                              12440.0  1103127.1      886.8
Process Creation                               126.0     3293.4      261.4
Shell Scripts (8 concurrent)                     6.0     1577.0     2628.3
System Call Overhead                         15000.0   840080.5      560.1
                                                                 =========
     FINAL SCORE                                                       0.0


Which one is better for polarhome?
Dell has a faster CPU, better file transfer
HP handles threads faster