Jump to content


Photo

QNAP TurboNAS TS-470 Review Discusison

NAS

  • You cannot start a new topic
  • Please log in to reply
7 replies to this topic

#1 Brian

Brian

    SR Admin

  • Admin
  • 5,309 posts

Posted 28 January 2014 - 04:17 PM

 

The QNAP TS-470 is a 4-bay NAS designed for SMB use (its sibling the TS-470 Pro is aimed at SOHO environments). Under the hood, the TS-470 offers high performance, powered by a dual-core Intel 2.6GHz processor and 2GB of DDR3 RAM (which can be expanded up to 16GB). Four Gigabit LAN ports (two single onboard and one dual-port gigabit Ethernet PCI-Express card) allow for maximum bandwidth efficiency; additionally, the dual-port NIC can be swapped out with a 10GbE adapter, greatly increasing IOPS throughput. With a mind toward scalability, the TS-470 offers a centralized storage solution for applications and environments with ever-growing storage needs, while also providing a full range of management applications and features through the QTS 4.0 operating system. Emphasis on data security, backup and retrieval, and antivirus also make the TS-470 an ideal choice for those who seek worry-free storage and safeguarding of increasing amounts of data.

 
QNAP TurboNAS TS-470 Review
 

Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#2 Makosuke

Makosuke

    Member

  • Member
  • 8 posts

Posted 29 January 2014 - 07:07 PM

There's a problem with this device (or, more generally, QNAP itself) that I discovered after buying into its little brother the TS-469L last month.  I'll explain it as a timeline:

 

  • They released version 4.0.5 of their very nice and easy to use firmware on December 20th, along with a beta of the upcoming 4.1.0.
  • It turns out that both the shipping version 4.0.5 and the beta had a bug in the AFP implementation.
  • This bug would cause a small percentage of files saved to (or in some cases just viewed on) the NAS to be corrupted.  If the file was copied to the NAS and not subsequently checked, there was in fact no warning whatsoever to the user--the copy completed and the filesize appeared approximately or exactly correct, but some of the data was replaced with junk, which you would discover the next time you went to look at the file.  Because of the quiet nature of the bug, that might be an instant save failure, or you might not notice for months if it was a drag-and-drop copy operation.

 

This is a very, very serious bug in a NAS.  Honestly, it's about as serious as a bug can possibly be--not only data loss, but data loss that could easily go un-noticed by the user.

 

This bug also completely broke MacOS Time Machine support, one of the advertised features.  A patch was eventually posted for that.

 

  • QNAP support became aware of this bug after a few days when people started complaining.  There are several threads in their forums as well.  They plan to eventually fix it in the final 4.1.0 release, according to support.
  • However--and this is the real problem--despite having a data-destructive bug in shipping firmware, posted on their website, they neither pulled that firmware nor put any kind of warning on it or the beta on the site.  It was just sitting there for unsuspecting users to download, install, and have data corrupted.
  • Eventually, about a month later, in late January, they finally pulled the 4.0.5 firmware and re-posted a version that presumably fixes the corruption bug.
  • The same build of the beta, with the bug, remains on their site, with no "known issues" warning whatsoever other than the standard disclaimer of "for testing only" and to be sure you've backed up your data.

 

Again, they had an acknowledged data-destructive bug in shipping firmware that they were aware of for somewhere in the vicinity of three weeks before they pulled and fixed the firmware publicly available on their site, and they neither before nor after issued any sort of warning to people who might have downloaded and installed it unless they contacted support (in which case they were told to use SMB because of the bug).

 

I like the hardware.  The software UI is fantastic.  The features are really good as well.  But after seeing that happen in real time, there's just no way I can trust the company to produce software I can trust with data of any importance whatsoever.  If they shipped data destructive firmware once and didn't issue any kind of warning for a month, who's to say the next time I install a firmware update (or buy a device with one already installed) it isn't going to have something equally egregious again?


Edited by Makosuke, 29 January 2014 - 07:10 PM.

#3 scaigs

scaigs

    Member

  • Member
  • 1 posts

Posted 29 January 2014 - 08:00 PM

Dear Brian,

 

A comment with your conclusion specially this phrase : The option to expand the unit with the 10 GbE adapter means that the unit is fully scalable in both terms of both raw storage space or speed.

 

Please note that the TS-x70 Series only have one expansion slot, so it means that either you have 10GbE or additional storage space but not both at the time.

 

I'm currently considering of acquiring the TS-870 with the 10GbE dual port but won't be able to expand the storage when the time comes, a hard decision and still hesitating to issue my purchase order based on this info.

 

Regards,


#4 Brian

Brian

    SR Admin

  • Admin
  • 5,309 posts

Posted 30 January 2014 - 09:49 AM

Dear Brian,

 

A comment with your conclusion specially this phrase : The option to expand the unit with the 10 GbE adapter means that the unit is fully scalable in both terms of both raw storage space or speed.

 

Please note that the TS-x70 Series only have one expansion slot, so it means that either you have 10GbE or additional storage space but not both at the time.

 

I'm currently considering of acquiring the TS-870 with the 10GbE dual port but won't be able to expand the storage when the time comes, a hard decision and still hesitating to issue my purchase order based on this info.

 

Regards,

 

Right, that's a fair point, as it is right now it's an either-or. I've emailed QNAP to find out if their upcoming tower expansion unit will have eSATA, which would make us right again ;) At least half right. 


Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#5 wlee

wlee

    Member

  • Member
  • 2 posts

Posted 15 April 2014 - 12:48 PM

Great review!

 

I'm wondering:

1. What's the drive configuration? I assume 4 HDD what's the raid group? and how many SSD and what's raid group used in SSD test?

2. Which iSCSI initiator was used? and what's the MPIO configuration?

3. What's CPU usage of NAS during testing?

4. Was iSCSI LUN file based?

 

Thanks.


Edited by wlee, 15 April 2014 - 01:45 PM.

#6 Kevin OBrien

Kevin OBrien

    StorageReview Editor

  • Admin
  • 1,449 posts

Posted 15 April 2014 - 03:29 PM

1. Both SSDs and HDDs were configured in RAID10

2. Can't recall the type off the top of my head, but it was the default one when creating iSCSI LUNs. We didn't use MPIO, but instead had two targets per IP.

3. Didn't poll that during the test.

4. 95% sure we used block based.. built the disk pool, created a volume for the CIFS shares, then added four 25GB iSCSI LUNs.


#7 wlee

wlee

    Member

  • Member
  • 2 posts

Posted 15 April 2014 - 03:55 PM

Thanks Brian.

 

2. NAS runs the target, initiator is at client side which connects to iSCSI target. Your sequential results of both 8K and 128K showed all four 1GbE NICs were utilized, almost saturated in 128K seq test. I'm curious about your client setup, used 10GbE NIC? How was client setup so that NAS' for 1Gbe NICs were utilized at the same time. iSCSI is my main interest, SMB setup is also good to know.


#8 Kevin OBrien

Kevin OBrien

    StorageReview Editor

  • Admin
  • 1,449 posts

Posted 15 April 2014 - 04:02 PM

When testing either 10GbE or 1GbE, we had either 2/2 port match or 4/4 port match on NAS/server. Each port was on its own subnet.






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users