Monday, April 30, 2012

VMware Fusion 4.1.2 update released today

About time...

From the update notes:
   Resolved an "out of memory" message when resuming some virtual machines.
   Resolved an intermittent printing problem when using Mac OS X Lion.
   Addressed an issue with copy and paste on newer Linux distributions.
   Fixed accessing files on shared folders using short names.
   Numerous Mac OS X compatibility improvements, including Developer ID.
   Resolved a power-off issue in Windows 2000.
   Updated version of McAfee VirusScan.

Sunday, April 29, 2012

Chrome vs Firefox 12.0, my memory usage tests


Test involves two windows, one with 10 tabs and another with 3 tabs, running on OSX 10.7.3


Before running either:



Wired Memory: 1309 MB
Active Memory: 3260 MB
Inactive Memory:         614 MB
Free Memory: 2546 MB
Real Mem Total (ps): 2460.141 MB


After starting Chrome:



Wired Memory: 1331 MB  (+22 MB)
Active Memory: 3990 MB  (+730 MB)
Inactive Memory:         640 MB  (+26 MB)
Free Memory: 1773 MB  (-773 MB)
Real Mem Total (ps): 3555.176 MB  (+1095.035 MB)

After viewing each tab (versus before start):

Wired Memory: 1323 MB  (+14)
Active Memory: 3981 MB  (+721)
Inactive Memory:         650 MB  (+36)
Free Memory: 1781 MB  (+781)
Real Mem Total (ps): 3563.785 MB  (+1103.644)



Now let's close Chrome (versus before starting Chrome):

Wired Memory: 1309 MB  
Active Memory: 3265 MB  (+5)
Inactive Memory:         622 MB   (+8)
Free Memory: 2530 MB  (-16)
Real Mem Total (ps): 2460.895 MB  (+0.754)

Time for Firefox:

Wired Memory: 1333 MB   (+24)
Active Memory: 3625 MB  (+360)  
Inactive Memory:         740 MB   (+118)
Free Memory: 2033 MB  (-497)
Real Mem Total (ps): 2958.316 MB  (+497.421)

And after viewing each tab (versus before start):

Wired Memory: 1324 MB   (+15)
Active Memory: 3639 MB (+374)
Inactive Memory:         748 MB  (+126)
Free Memory: 2018 MB  (-512)
Real Mem Total (ps): 2968.293 MB  (+507.398)

Now with Firefox closed  (versus before starting Firefox):

Wired Memory: 1310 MB  (+1)
Active Memory: 3279 MB  (+14)
Inactive Memory:         636 MB  (+14)
Free Memory: 2505 MB  (-25)
Real Mem Total (ps): 2478.117 MB  (+17.222)

So that is Chrome using more memory versus Firefox, and here's how much:

Wired Memory: -1 MB  
Active Memory: +347 MB  (92.7% increase over FF)
Inactive Memory:         -98 MB  (77% decrease over FF)
Free Memory: +269 MB  (52.5% increase over FF)
Real Mem Total (ps): +596.246 MB  (117% increase over FF)


Pass #2 for Chrome, before running:

Wired Memory: 1345 MB
Active Memory: 3155 MB
Inactive Memory:         606 MB
Free Memory: 2676 MB
Real Mem Total (ps): 2393.332 MB

After running:

Wired Memory: 1347 MB  (+2)
Active Memory: 3819 MB  (+664)
Inactive Memory:         621 MB  (+15)
Free Memory: 2016 MB  (-660)
Real Mem Total (ps): 3414.230 MB  (+1020.898)

Again we see Chrome using more memory than Firefox.

More testing to come in the future, but this is pretty disappointing for Chrome.


memory - Is there a Mac OS X Terminal version of the "free" command in Linux systems? - Apple - Stack Exchange

memory - Is there a Mac OS X Terminal version of the "free" command in Linux systems? - Apple - Stack Exchange

Sans Digital 4-port eSATA PCIe Host bus adapter (HA-DAT-4ESPCIE) Review

Today I'm reviewing the Sans Digital 4ESPCIE four eSATA Ports PCI-Express (x8) Host Adapter . The Sans Digital 4ESPCIE HA-DAT-4ESPCIE uses a Silicon Image 3124  (Sil3124) chipset, and has 4 pairs of LED pins, one for each eSATA port.  It has one jumper set for Enabling (Pin 1+2) and Disabling (Pin 2+3) the Bios.  On the back is a sticker labeled "ESATAPCI8 , 1211081"

This controller will be connected to a Sans Digital TowerRAID TR8M+B - 8 Bay eSATA JBOD Performance Tower with 6G PCIe Card (Black), a very capable JBOD enclosure I have been running for 2 months now.

With the BIOS enabled, at BIOS start only one drive shows, drive 1, in my Sans Digital SATA enclosure.  However, when boot into Ubuntu Linux, all drives activate fine through AHCI kernel driver.

When booted into Gentoo however, only one drive showed up.

# lspci:
03:00.0 RAID bus controller: Silicon Image, Inc. SiI 3124 PCI-X Serial ATA Controller (rev 02)


The PEXESAT32 I was running worked fine with my kernel, must need a different driver:

# make menuconfig
 Device Drivers --->
 (*) Serial ATA and Parallel ATA drivers  --->
    Silicon Image 3124/3132 SATA support   


# make
scripts/kconfig/conf --silentoldconfig Kconfig
  CHK     include/linux/version.h
  CHK     include/generated/utsrelease.h
  CALL    scripts/checksyscalls.sh
  CHK     include/generated/compile.h
  CC [M]  drivers/ata/sata_sil24.o
Kernel: arch/x86/boot/bzImage is ready  (#3)
  Building modules, stage 2.
  MODPOST 160 modules
  CC      drivers/ata/sata_sil24.mod.o
  LD [M]  drivers/ata/sata_sil24.ko

# modprobe sata_sil24


[ 2036.028575] sata_sil24 0000:03:00.0: version 1.1
[ 2036.028585] sata_sil24 0000:03:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
[ 2036.028770] sata_sil24 0000:03:00.0: Applying completion IRQ loss on PCI-X errata fix
[ 2036.029191] scsi11 : sata_sil24
[ 2036.029254] scsi12 : sata_sil24
[ 2036.029421] scsi13 : sata_sil24
[ 2036.029474] scsi14 : sata_sil24
[ 2036.029505] ata11: SATA max UDMA/100 host m128@0xf78ffc00 port 0xf78f0000 irq 16
[ 2036.029508] ata12: SATA max UDMA/100 host m128@0xf78ffc00 port 0xf78f2000 irq 16
[ 2036.029510] ata13: SATA max UDMA/100 host m128@0xf78ffc00 port 0xf78f4000 irq 16
[ 2036.029512] ata14: SATA max UDMA/100 host m128@0xf78ffc00 port 0xf78f6000 irq 16
[ 2038.155629] ata11: SATA link up 3.0 Gbps (SStatus 123 SControl 0)
[ 2038.155986] ata11.15: Port Multiplier 1.1, 0x1095:0x3726 r23, 6 ports, feat 0x1/0x9
[ 2038.157041] ata11.00: hard resetting link
[ 2038.472072] ata11.00: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
[ 2038.472104] ata11.01: hard resetting link
[ 2038.787328] ata11.01: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 2038.787359] ata11.02: hard resetting link
[ 2039.102571] ata11.02: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 2039.102603] ata11.03: hard resetting link
[ 2039.428778] ata11.03: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 2039.428810] ata11.04: hard resetting link
[ 2039.733186] ata11.04: SATA link down (SStatus 0 SControl 320)
[ 2039.733228] ata11.05: hard resetting link
[ 2040.037405] ata11.05: SATA link up 1.5 Gbps (SStatus 113 SControl 320)
[ 2040.041678] ata11.00: ATA-8: WDC WD30EZRX-00MMMB0, 80.00A80, max UDMA/133
[ 2040.041683] ata11.00: 5860533168 sectors, multi 16: LBA48 NCQ (depth 31/32)
[ 2040.045237] ata11.00: configured for UDMA/100
[ 2040.046906] ata11.01: ATA-8: Hitachi HDS5C3030ALA630, MEAOA580, max UDMA/133
[ 2040.046911] ata11.01: 5860533168 sectors, multi 0: LBA48 NCQ (depth 31/32)
[ 2040.048641] ata11.01: configured for UDMA/100
[ 2040.053384] ata11.02: ATA-8: WDC WD30EZRX-00MMMB0, 80.00A80, max UDMA/133
[ 2040.053389] ata11.02: 5860533168 sectors, multi 0: LBA48 NCQ (depth 31/32)
[ 2040.058378] ata11.02: configured for UDMA/100
[ 2040.059568] ata11.03: ATA-8: ST31500541AS, CC34, max UDMA/133
[ 2040.059572] ata11.03: 2930277168 sectors, multi 0: LBA48 NCQ (depth 31/32)
[ 2040.060959] ata11.03: configured for UDMA/100
[ 2040.061091] ata11: EH complete
...

Perfect.


 # zpool status
  pool: vault
 state: ONLINE
 scan: scrub repaired 0 in 23h37m with 0 errors on Wed Apr 18 20:02:24 2012
config:

NAME                                            STATE     READ WRITE CKSUM
vault                                           ONLINE       0     0     0
 raidz2-0                                      ONLINE       0     0     0
   ata-Hitachi_HDS5C3030ALA630_MJ1311YNG35RYA  ONLINE       0     0     0
   ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0239674    ONLINE       0     0     0
   ata-Hitachi_HDS5C3030ALA630_MJ1311YNG62YAA  ONLINE       0     0     0
   ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0065897    ONLINE       0     0     0
   ata-Hitachi_HDS5C3030ALA630_MJ1311YNG3B2EA  ONLINE       0     0     0
   ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0059589    ONLINE       0     0     0

errors: No known data errors

# zfs mount vault

As a note, I have the array connected to port 1 and port 3 on the card.


# dd if=VMware-server-2.0.2-203138.exe of=/dev/null bs=1M
507+1 records in
507+1 records out
532132088 bytes (532 MB) copied, 2.98222 s, 178 MB/s

# dd if=VMware-server-2.0.2-203138.i386.tar.gz of=/dev/null bs=1M
482+1 records in
482+1 records out
506047036 bytes (506 MB) copied, 2.53113 s, 200 MB/s

Let's do a zpool scrub, which would only reach ~80M/s using the PEXESAT32 card:

# zpool status
  pool: vault
 state: ONLINE
 scan: scrub in progress since Mon Apr 23 14:59:59 2012
    20.8G scanned out of 6.46T at 155M/s, 12h4m to go
    0 repaired, 0.31% done
config:

NAME                                            STATE     READ WRITE CKSUM
vault                                           ONLINE       0     0     0
 raidz2-0                                      ONLINE       0     0     0
   ata-Hitachi_HDS5C3030ALA630_MJ1311YNG35RYA  ONLINE       0     0     0
   ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0239674    ONLINE       0     0     0
   ata-Hitachi_HDS5C3030ALA630_MJ1311YNG62YAA  ONLINE       0     0     0
   ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0065897    ONLINE       0     0     0
   ata-Hitachi_HDS5C3030ALA630_MJ1311YNG3B2EA  ONLINE       0     0     0
   ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0059589    ONLINE       0     0     0

errors: No known data errors

Very nice improvement, and it keeps getting better:

# zpool status
  pool: vault
 state: ONLINE
 scan: scrub in progress since Mon Apr 23 14:59:59 2012
    89.1G scanned out of 6.46T at 186M/s, 9h59m to go

# zpool status
  pool: vault
 state: ONLINE
 scan: scrub in progress since Mon Apr 23 14:59:59 2012
    566G scanned out of 6.46T at 191M/s, 9h1m to go


After running for 6 days, system is completely stable and performance is much better compared to my prior eSATA setup.


This is certainly a great inexpensive alternative to a more expensive SAS deployment.

More test results to come...


Wednesday, April 18, 2012

DroboS Support for eSATA... or rather lack of. Yikes, this is quite a read...

Drobo Support:
Same specific cards noted as problems:

We are working with the vendors of a number of cards to resolve compatibility issues. Until these are resolved, we recommend not using the following cards:
Seritek 2S2E-E & SeriTek/2ME4-E, SeriTek/e6G
MAXPower eSATA 6G PCIe 2.0 Controller Card
Sonnett Tempo E4P
Sonnet Tempo ProExpressCard /34
Sonnet Tempo X4P
Sonnet Qio eSATA ports
All RocketRAID cards
StarTech 1 Port PCI-Express eSATA Card� (P/N: PEXSATA1)