Areca Firmware Update V1.47 Dated 19th August

Printable View

08-25-2009, 01:54 AM
eternal_fantasy

Areca Firmware Update V1.47 Dated 19th August

Release Notes:

********* Change Log For V1.47 Firmware

2009-1-13
1 Fix R1+0 128K sequential read hang when stripe size is 4K, and odd number of
HDD is used
2 Fix multiple volumeset rebuilding sequence problem
2009-1-15
1 Fix SNMP bulk request and response greater than 1500 bytes. Support UDP
fragmentation.
2009-2-16
1 If rebuilding+migrating, we need to use MIGRATING type init routine
2009-2-18
1 Add DHCP get in inifinte loop if
(A) DHCP is enabled
(B) IP setting is 0.0.0.0
2009-2-24
1 Add HDD APM support
(A) Add HDD Power Management item
(B) Stagger power on and standby functions are moved to HDD Power Management
(C) Add APM Low Power Idle (2/3/4/5/6/7 minutes)
(D) Add APM Low RPM Standby (10/20/30/40/50/60 minutes)
(E) Original Standby time is added with max of (low power idle and low RPM
stanby) time. And HDD enter spin down state.
(F) Only Hitachi HDDs are supported
(G) Seagate and WDC hdd report that NO APM support
(H) SAMSUNG report APM support and function is incorrect, we exclude
Samsung HDD
(I) If HDD is in Low RPM state, it may take longer than 8 seconds to spin up
2009-2-25
1 Set raidTimeAccess during GetFreeRaidSet to prevent raidset hdd spun down
immediately after raid created
2009-2-27
1 check US daylight saving problem
2 ARC1222 add R5060 support
2009-3-3
1 Fix GuiExpandRaidSet
2009-3-6
1 Fix ARC1201/1200, after reboot, the controller missing
2009-3-31
1 Fix SATA raid controller seagate HDD error handling
2009-4-2
1 Fix ARC1200+supermicro M/B problem
2 ARC1680, modify SendAbortTask
2009-4-7
1 Improve Error Handling
2 Add scdl_devmap_clear_reset_state
2009-4-10
1 Fix ARC1201/1200 shutdown problem
2 Porting of PCI RAID card to support upto 4MB request for normal xfer
3 Driver Modification
2009-4-16
1 ARC1201:reading of smart hdd temperature, if hdd pug/unplug may cause
smart command to fail and may cause the hdd temp to be disabled.
2009-4-29
1 ARC1680:Patch PMC-SIERRA expander for seagate hdd SATA_CONT problem,
2 ARC1680:Patch Vitesse expander for seagate hdd SATA_CONT problem,
3 NO_READ_AHEAD_1M
4 PCI/E-RAID:Add NEW_DMA_HANDLING to support up to 4MB xfer per request
5 ARC1680:use link reset to abort timeout command
6 ARC1680:Fix seagate update firmware problem
7 Fix SetWriteMask for RAID6
2009-5-8
1 Improve AJA performance
2 TYAN S5375 M/B (Intel5100 chip set), Intel specification update:cannot use
read completion coalescing
2009-5-14
1 NEW_DIRTYLINE_HANDLING
2009-5-15
1 Add SUPPORT_WRITE_THROTTLING
2009-5-21
1 ARC1680:fix JBOD mode scsi id assignment error
2 ARC1680/1212:CPU fan not detected by default
2009-5-25
1 Add PING_PONG_PCIDMA
2 Improve SAFARI support, but not solved
2009-5-27
1 Improve slow volume performance for ARC1680 by adding
LOCAL_CACHE_LOCK
LOCAL_GLOBAL
2 ARC1680:transport changed to V4.7.2.0
2009-6-2
1 Fix Opensolaris+ZFS to add device to mirror set in JBOD or passthrough mode
2 ARC1680 BIOS:add 3 minutes timeout
2009-6-19
1 Fix ARC1201: check timeout for NON-Initialized channel
2 Do not rebuild all failed raidset:
3 Add I2C_TIMEOUT_SUPPORT for ARC1680
4 Fix SAFARI 4.0 HTTP support, change ONCE random number
2009-6-25
1 V1.47 2009-06-25 for All PCI model
2009-7-2
1 Remove PING_PONG_PCIDMA support
2 Fix ARCHTTP problem for V1.47 if no ethernet port adapter is used
ARC1110/1120/1200/1210/1220/1211/1221

:up:
08-25-2009, 06:52 AM
F.E.A.R.

Oooold :D

I´m testing 1.47 just in this moment.
My conclusion - 1.46 is faster.

Setup: 2x Acard 9010, Areca ARC-1210, XP Pro x86, Vista Ultimate x64, W7 Ultimate x64 RTM
Red ist advantage for 1.46 - blue is advantage for 1.47

XP x86 - Stripesize 4k

create 0%
read +59%
copy +13%

XP x86 - Stripesize 128k

create 0%
read 0%
copy -4%

vista x64 - Stripesize 4k

create 0%
read -2%
copy -3,5%

vista x64 - Stripesize 128k

create 0%
read +31%
copy +13%

w7 x64 - Stripesize 4k

create 0%
read +10%
copy +5%

w7 x64 - Stripesize 128k

create 0%
read +13%
copy +6%
08-25-2009, 07:22 PM
tommyshango

Is there any worry about updating firmware and losing data? I know there usually isnt but I am just scared to try it.

Also any idea if moving around cables would affect an array? I have quite a few cables and added more drives and wanted to clean up the appearance by putting them in a more logical order. Any worries?

Thanks
08-25-2009, 07:35 PM
WangChung

The Areca controllers are supposed to auto detect which drive is which no matter what location you put the cable in before and after disconnection. Haven't given it a shot but that's what the documentation says. :)
08-26-2009, 01:49 PM
F.E.A.R.

Can anyone confirm my results?
08-26-2009, 03:36 PM
WangChung

Quote:

Originally Posted by F.E.A.R.

Can anyone confirm my results?

Which benchmark(s) did you use?
08-26-2009, 06:18 PM
F.E.A.R.

FC-Test
08-27-2009, 07:29 AM
alfaunits

Quote:

Originally Posted by WangChung

The Areca controllers are supposed to auto detect which drive is which no matter what location you put the cable in before and after disconnection. Haven't given it a shot but that's what the documentation says. :)

NEVER count on this! If it were only me, I would say "bad luck", but too many people have seen this not to work.

Heck, Areca looses drives if they remain in the same port even :(
08-27-2009, 10:10 AM
WangChung

^^ Yeah, I've never tempted fate. But the documentation says it supports it. All my drives are arranged top to bottom in the port number they belong to. Actually, next paycheck when I get my new drive I'll give it a shot and report back. :)
08-27-2009, 12:18 PM
alfaunits

It may work - but as I said, Areca looses drives even when you don't change ports. The few times I tried changing them - it "magically" got several incomplete volumes instread of one.
So go figure ;) I'm not the only one.
08-27-2009, 08:52 PM
Nizzen

I did not have any problems with disks dropping out with my Areca 1680ix. Have used it for a year now. I tested many kinds of disks with great suksess. Seagate 15k.5 sas, Fujitsu 15k sas, WD 750 se sata drives, samsung 1tb f1 drives, seagate 15k.4 sas drives, samsung 500gb drives. And some old maxtor maxline III drives :p

I change ports many times with many types of raid, and the areca allways finds the array. So not all have problems. Just they who have problems shouts, but for the many others who not have any problems shut up.

This is just my oppinion.
08-27-2009, 10:03 PM
alfaunits

You shouldn't shout "yes this work!" if it should work.
I never said it does not work for everyone, do not put words in my posts ;)
It is enough that several people have reported problems that this should be considered serious by Areca - it wasn't.
Good for you that it worked! Guess what??? It didn't work for everyone! That's what counts.
08-27-2009, 10:10 PM
F.E.A.R.

Quote:

Originally Posted by Nizzen

I change ports many times with many types of raid, and the areca allways finds the array. So not all have problems. .

dito

Is 1.47 a problem-fw?
08-28-2009, 07:21 AM
henk53

We're going to test 1.47 very soon too on a 1680ix-8, probably with 8xX25-E. Tests will be done on Linux with XFS using xdd and easyco. I'll report back later (hopefully sooner than later).
08-28-2009, 07:23 AM
F.E.A.R.

Only Linux?
08-28-2009, 07:24 AM
henk53

Btw, about this change:

ARC1680:transport changed to V4.7.2.0

Does anyone has further details about what changed in the transport part? Can't find any details about this on the Intel site.
08-28-2009, 07:27 AM
F.E.A.R.

Ask Areca-Support.
08-28-2009, 08:17 AM
henk53

Quote:

Originally Posted by F.E.A.R.

Only Linux?

Yes, all of our servers, where we use these setups, are running Linux. Incidentally, most engineers are running Linux on their desktops too and the engineers which are not running Linux run OS X.

If we had a Windows license available somewhere I would be happy to do a quick test. On the other hand, I don't have a lot of experience with Windows really, so a quick test might not even be possible as I have to learn how to operate Windows (although this probably shouldn't be that difficult I guess...).
08-30-2009, 09:41 PM
dhc0329

I was using v1.45 with Velocirapters x 2 (raid 0). I coudn't use v1.46 because it was rather unstable. v1.47 seems to be stable but performace wise it is slightly inferior to v1.45 on HD Tune (read access test).
09-01-2009, 09:32 AM
F.E.A.R.

I retry the tests (1.46 vs. 1.47)...
Conclusion - 1.46 runs faster on the ARC-1210 an d 1.47 faster on the ARC-1261ML.
Why these big differences???
09-03-2009, 06:17 AM
Mechromancer

I see no enhancements for SSDs. I'll stick with 1.46 for my RAID-0 setup. My IOP333 doesn't have to do much on that setup anyway.
09-03-2009, 03:52 PM
F.E.A.R.

Try FC-Test and you see a difference
09-05-2009, 04:29 AM
Mechromancer

Quote:

Originally Posted by F.E.A.R.

Try FC-Test and you see a difference

You mean you see a difference with SSDs?
09-05-2009, 05:47 AM
F.E.A.R.

Yepp... http://www.xtremesystems.org/forums/...57&postcount=2
09-05-2009, 07:33 AM
SteveRo

Quote:

Originally Posted by F.E.A.R.

Yepp... http://www.xtremesystems.org/forums/...57&postcount=2

Fear - what kind of improvement did you see on 1231ML or do you have the 1261?
09-05-2009, 07:50 AM
F.E.A.R.

I´m using a ARC-1210 and 1261ML
I get the best results with 1.47 on the ARC-1261ML + Vista.
The real improvement is ~ 5 sec. by copying of 18.200 small files (5,6GB) inside the array.

XP x86 SP3
4k -100% random - 100% write - 4k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_799xpx...64k12695j9.jpg http://www.abload.de/thumb/hc_810xpx...74k126m6ff.jpg

XP x86 SP3
4k -100% random - 100% write - 128k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_801xpx...6128k1b3bc.jpg http://www.abload.de/thumb/hc_813xpx...7128k152c4.jpg

Vista x64 SP2
4k -100% random - 100% write - 4k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_798vis...1.464kz6os.jpg http://www.abload.de/thumb/hc_809vis...1.474km5me.jpg

Vista x64 SP2
4k -100% random - 100% write - 128k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_795vis...1.4612y657.jpg http://www.abload.de/thumb/hc_806vis...1.4712215v.jpg

WIN 7 x64 RTM (final version)
4k -100% random - 100% write - 128k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_802w7x...6128k1qgmw.jpg http://www.abload.de/thumb/hc_814w7x...7128k1ehs9.jpg

But the ARC-1210 is faster with 1.46 :confused:

XP x86 SP3
4k -100% random - 100% write - 4k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_784xpx86fw1.464koe9v.jpg http://www.abload.de/thumb/hc_778xpx86fw1.474ksfnd.jpg

XP x86 SP3
4k -100% random - 100% write - 128k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_786xpx...46128kpfh7.jpg http://www.abload.de/thumb/hc_780xpx...47128kqhg5.jpg

Vista x64 SP2
4k -100% random - 100% write - 4k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_772vis...1.464k3i1r.jpg http://www.abload.de/thumb/hc_777vis...1.474kqi50.jpg

Vista x64 SP2
4k -100% random - 100% write - 128k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_770vis...1.46125fl7.jpg http://www.abload.de/thumb/hc_775vis...1.4712bfoq.jpg

WIN 7 x64 RTM (final version)
4k -100% random - 100% write - 128k stripesize

fw1.46 - fw1.47

http://www.abload.de/thumb/hc_787w7x...46128kov81.jpg http://www.abload.de/thumb/hc_781w7x...47128kp76n.jpg
09-05-2009, 10:50 AM
SteveRo

FEAR - ok understand, I was using all the usual benchmarks and not seeing any material difference.
I will stay on 1.47 - any change in the best strip size for the acards?