Talk:NetApp FAS
![]() | This article is rated C-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
Aggregate Limits & capacity overhead
editFAS2020 has been extended to 16 TB in the most recent software release 7.3.1
The capacity overhead is "working as designed" as it provides additional checksumming for enhanced security as well as file system overhead similar to ntfs, ext3 or others. This is a common practice for most enterprise storage vendors "right sizing" hard disks to enable different hard disks from different vendors in the same raid set. —Preceding unsigned comment added by 87.139.22.228 (talk) 14:55, 27 January 2009 (UTC)
x86
edit"although based on x86" - true to some extent, but not 100% accurate. FAS200 series appliances use MIPS processors - these are not x86. New units (FAS6080 for example), use x64 based AMD chips...
- I've removed that. --Kubanczyk (talk) 06:53, 12 May 2008 (UTC)
NVRAM vs NVMEM
editThe article claims that all filers have separate dedicated NVRAM. This has not been the case for several of the machines and instead they use NVMEM where a portion of the system main RAM maintains power even though the CPU and the rest of the motherboard has lost power. Examples incude the FAS250, 270, 2040, 2220, 2240, and the entire 32xx range. — Preceding unsigned comment added by 198.95.226.40 (talk) 20:59, 19 July 2012 (UTC)
Memory / RAM table is incorrect
editFor example, the 6280 does not have 192GB of ram but instead has half of that. The table is mixing the values for HA pairs with single controllers. — Preceding unsigned comment added by 198.95.226.40 (talk) 21:02, 19 July 2012 (UTC)
I agree, this is not quite right. The CPU counts appear to be per controller, but the memory is shown for the total of an HA pair system. EG: FAS3240 has 1x L5410 CPU per controller and 8GB ram per controller. FAS3220, also has 1x L5410 cpu and 12GB ram per controller.. — Preceding unsigned comment added by 202.167.15.164 (talk) 02:50, 7 February 2017 (UTC)
Authentication
editNothing in the article about how authentication and authorization interoperate with various directories. 1.44.43.140 (talk) 12:04, 12 May 2014 (UTC)
That's a function of ONTAP, and would belong in the Data ONTAP article. Guy Harris (talk) 02:16, 28 June 2018 (UTC)
Neutrality
editThe article is pretty modest, it doesn't make a big point of bringing NAS appliances to a larger audience, nor the immense benefits of introducing nvram (closing the gap sun left at prestoserve), nor does it make a large point of the, well, basically the invention of copy on write filesystems. An advertisement could sound pretty different; anyone with reasonable industry experience should think this is really neutral. If anything, the article lacks in information about the issues the by now rather old architecture started to show.
I'm pretty sure if I pull up articles for OSS clones like Openfiler or FreeNAS those will be much more of an advertisement. 2001:A60:16A8:DE01:F166:9635:CA24:A655 (talk) 20:24, 27 July 2015 (UTC)
New title needed - NetApp storage appliance?
edit"Filer" is no longer being used now that block protocols are also used, so the article should have a better title, rather than having the somewhat weird lede it currently has. "NetApp storage appliance" might work; NetApp appear to use it in some places.
The article should indicate that "filer" was the term originally used, but shouldn't continue to use "filer" in the rest of the article. Guy Harris (talk) 02:16, 28 June 2018 (UTC)
Someone named Drmies keeps deleting the Model History chart section.
editCan we stop this person? He has already removed it twice, and two of us had to undo the removal.
The chart has been in the article for 11 years! It serves a useful purpose, and is not documented anywhere else outside of Wikipedia. — Preceding unsigned comment added by 216.240.30.11 (talk) 01:21, 4 June 2019 (UTC)
Hi, I also today missed the FAS hardware table and looked into drmies's arguments against that table: "all of this we can get from the NetApp website" - this is plainly wrong, addditionaly there are no other sources for most of the old models "this excessive and unencyclopedic detail" - this is usual and necessary in articles which document characteristics of evolution of hardware product lines, f.e. https://en.wikipedia.org/wiki/List_of_Intel_Xeon_microprocessors , https://en.wikipedia.org/wiki/Zen_(microarchitecture) "that lacks secondary sources and thus can make no claim to being relevant" It is also wrong that there are no secondary sources. F.e. Customers can see a models spec at boot up, compare to information delivered by manufacturers SE and contribute those to the community. The relevance rises with time. The older the devices, the more difficult to find those informations somewhere. In many cases of older models there are no open accessible sources anymore available except wiki. "So it's pure original research? " - This question puzzles me, as it indicates drmies does judge the tables information and relevance without knowning how this data was determined. — Preceding unsigned comment added by 91.58.4.136 (talk) 15:53, 5 July 2019 (UTC)
The Link to James Lau is wrong
editIt points to the wrong person. — Preceding unsigned comment added by 2003:f6:3f1e:7e00:8405:ec23:329b:de3a (talk) 11:36, 17 July 2020 (UTC)
- I removed the link. Guy Harris (talk) 14:45, 17 July 2020 (UTC)