Oracle is putting a heavy focus on Sun's 7000, tape and flash storage products, implying the rest of the bought-in storage line is not so well-regarded and may be under threat. Larry E and the new exec team at the big O have spoken and the storage winners and losers are becoming clear. In the winners' enclosure are the …
If they want to replace the 99xx arrays with a beefed-up 7xxx they'll have to implement a proper replication product in them that can match up to Hitachi's universal replicator. The incremental snapshot-and-send ZFS model is adequate for remote backups, but not for enterprise-class disaster recovery.
That would be DataGuard
Welcome to Oracle - the database does the clever stuff.
7000 ZFS and NetApp
Did SUN and NetApp settle up on ZFS already or it is still open?
@ 7000 ZFS and NetApp
Some people have short memories.
Did NetApp ever settle the patent infringement with StorageTek over the STK Shared Virtual Array (Iceberg) ? The first product by a long chalk to implement a virtual storage array based on "no update in place". A time when NetApps employees were in nappies.
Oh I forgot, Sun, I mean Oracle, now own that intellectual property ....
Sun Storage overall...
Well, even though the 25x0 storage seems to be much maligned, we've actually been quite happy with it.
I know they're dropping the j4x00 JBOD's - which we were actually starting to really like - this year (not to be confused with the j4400's they sell for 7000 series integration).
The 7000 series are fantastic boxes for what they are, but if you're wanting to run VDI or OLTP type stuff from them, you're still better off running in RAID 0+1 rather than one of the RAID-Z's. Physics still gets in the way.
We'll probably put in 4 to 6 of the 7410 arrays this year, and maybe a dozen 2540's. The 7310's and 7410's aren't quite as refined as a NetApp (we have a FAS3170, and we'll probably end up with a 2040 or two this year, so it's not like we've entirely drunk the Sun Kool-Aid) but if you need/want all the additional stuff Sun just includes for "free," there is no comparison in price. It's not even close.
I wish they wouldn't drop their low end though. I hate to think we're gonna be stuck using Dell (especially storage - the R610 and R710 aren't THAT bad once you get past iDRAC6 Enterprise's issues.) The AX4-5f is a mess compared to a 2540, and significantly more expensive.
would like to see some SPC numbers from the 7000, the 6000 posted numbers/pricing isn't impressive.
If you look at all the mess the other players have, (e.g. EMC with Celerra, Clariion, Symmetrix), I think Oracle does the right thing.
ZFS (in combination with Solaris) already does NFS, CIFS, iSCSI, FC, DeDup, Compression, Async-Replication). It needs to be polished/improved a little bit to be really at the high-end level. But absolutely doable in 2-3 years time.
It does NOT do async replication. Don't confuse snapshots with async. These arrays do remote backups, not replication. Ask Sun how you perform a reverse resync. Then work out how long it takes to copy , say, 1TB across your network...
If you do a ZFS snapshot, and backup the snapshot remotely, then only the bits that changed will be transmitted. Not all data will be transmitted. This is old ZFS functionality.
But with dedup it works like this. If you do a ZFS snapshot and backup the snapshot remotely, and if you transmit data blocks that already exist on the remote server, then ZFS will only send references to the data blocks. The actual data will not be sent. Only new data blocks that dont exist remotely, will be sent. Everything else will be sent as references, pointing to the data that already exist on remote server. There are potentially enormous savings in bandwidth here. Combine this with compression too, if you wish. :o)
offtopic, but what issues do you have with the iDRAC6's?
my issue is getting an SSL certificate to load.
- Vid Hubble 'scope snaps 200,000-ton chunky crumble conundrum
- Bugger the jetpack, where's my 21st-century Psion?
- Google offers up its own Googlers in cloud channel chumship trawl
- Windows 8.1 Update 1 spewed online a MONTH early – by Microsoft
- Interview Global Warming IS REAL, argues sceptic mathematician - it just isn't THERMAGEDDON