killerpub

回復 30# icarus-c


    要去細嘗一下 btrfs wiki 啦

一支手指用 subvolume 可以好多 root
即係好多個 Linux lives in a single partition

cpu 夠應付 compress 就行了

投資 effort 在 Linux ZFS, 個人上唔值得。
至於 FreeBSD 都係得個追字，即係夢。

t1066

回復 31# killerpub

正在使用 FreeBSD + zfs v28 的路過

t1066

基本上，若不使用dedup或encrypt等功能的話，cpu的要求並不太高，當然ram是越多越好。

computer1975

我一般在公司既 production server 都係用 XFS 同 EXT3 On CentOS
最近測試轉左d  EXT3 上 EXT4 ...
frank7lan 發表於 2011-9-1 12:17

終於找到有ching用sgi XFS

lazyfai

我都用咗好多年啦。不過之前炒過一次大大鑊就記得要成日backup了。

終於找到有ching用sgi XFS
computer1975 發表於 2011-9-3 20:57

ipsec

基本上，若不使用dedup或encrypt等功能的話，cpu的要求並不太高，當然ram是越多越好。 ...
t1066 發表於 2011-9-3 11:29

    睇過一些FreeNAS+ZFS 的報道、最好每1TB disk space 要 1GB RAM.

t1066

睇過一些FreeNAS+ZFS 的報道、最好每1TB disk space 要 1GB RAM.
ipsec 發表於 2011-9-7 22:11

由於zfs可以用SSD作為cache drive，所以未必需要1G ram對1T data，就好像以下：

https://rcbi.rochester.edu/users ... nd_OpenSolaris.html

不過要注意的是，reboot後cache drive會重新運作，失去之前的cache。所以比較適合長期運行的server.

ipsec

回復 37# t1066


    I think the quoted example is a extremely case, the writer use 13 x 2TB hard disks on his test lab which is not a common use for SOHO/Home user.

http://doc.freenas.org/index.php/Hardware_Requirements

"RAM
The best way to get the most out of your FreeNAS™ hardware is to install as much RAM as possible. If your RAM is limited, consider using UFS until you can afford better hardware. ZFS typically requires a minimum of 6 GB of RAM in order to provide good performance; in practical terms (what you can actually install), this means that the minimum is really 8 GB. The more RAM, the better the performance, and the Forums provide anecdotal evidence from users on how much performance is gained by adding more RAM. For systems with large disk capacity (greater than 6 TB), a general rule of thumb is 1GB of RAM for every 1TB of storage. "

"Storage Disks and Controllers
.....For example, moving the ZIL (ZFS Intent Log) to a dedicated SSD only helps performance if you have synchronous writes, like a database server. SSD cache devices only help if your working set is larger than system RAM, but small enough that a significant percentage of it will fit on the SSD. "

t1066

回復 38# ipsec

I would refer to the following article:

http://constantin.glez.de/blog/2 ... esystem-performance

#4: Use SSDs to Improve Read Performance

If you can't add any more RAM to your server (or if your purchasing department won't allow you), the next best way to increase read performance is to add solid state disks (aka flash memory) as a level 2 ARC cache (L2ARC) to your system.

You can easily configure them with the zpool(1M) command, read the "Cache devices" section of its man-page.

SSDs can deliver two orders of magnitude better IOPS than traditional harddisks, and they're much cheaper on a per-GB basis than RAM.
They form an excellent layer of cache between the ZFS RAM-based ARC and the actual disk storage.

You don't need to observe any reliability requirements when configuring L2ARC devices: If they fail, no data is lost because it can always be retrieved from disk.

This means that L2ARC devices can be cheap, but before you start putting USB sticks into your server, you should make sure they deliver a good performance benefit over your rotating disks :).

SSDs come in various sizes: From drop-in-replacements for existing SATA disks in the range of 32GB to the Oracle Sun F20 PCI card with 96GB of flash and built-in SAS controllers (which is one of the secrets behind Oracle Exadata V2's breakthrough performance), to the mighty fast Oracle Sun F5100 flash array (which is the secret behind Oracle's current TPC-C and other world records) with a whopping 1.96TB of pure flash memory and over a million IOPS. Nice!

And since the dedup table is stored in the ZFS ARC and consequently spills off into the L2ARC if available, using SSDs as cache devices will also benefit deduplication performance.

As you can see, an SSD or a good USB drive could be used as poor man's RAM. Moreover, in a Home/SOHO setting, I would guess a typical working set will be less than 100G which can easily fit into a 128G SSD.

ipsec

回復 39# t1066

From the same article, the writer mentioned "Add Enough RAM" & "Add More RAM" on top of the #1 & #2 places about read/write improvement and "Use of SSD for data caching only if you can't add any more RAM to your server.".

The writer also mentioned "Having enough RAM will benefit all of your reads, no matter if they're random or sequential, just because they'll be easier for ZFS to find on your disks, so make sure you have at least n/1000 + 1 GB of RAM, where n is the number of GB in your storage pool.".

It depends the storage size requirement for the installation, whether they should go to take the benefit of SSD as a second level cache. For the HOME/SOHO users, it is not difficult to build a PC/Server with 16G or even more than 32G RAM memory now day. According to the writer's rule, "n/1000 + 1 GB of RAM", that means the PC/Server can build a 15TB or 31TB more disk storage pool to be used. Some of them may add a hardware RAID card solution instead of building software RAID-Z environment.

As an installation have to use SSD as cache, their disk storage requirements are most likely greater than 31TB. Such installation, the user would like to concern on the data integrity/security/backup/recover support rather than the read/write performance by purchasing ORACLE/IBM/EMC SAN solution in order to make their invaluable data safety. Of course, data read/write performance is a value add factor.