Tag Archive for sparc

Sun SPARC T3 Servers

Oracle announced their new line of Sun SPARC T3 powered servers at Oracle Openworld 2010. The SPARC T3 processor includes several improvements on T2 and T2+ processors including:

T2 / T2+T3
65 nm manufacturing process40 nm manufacturing process
4MB L2 Cache6MB L2 Cache
8 Cores (8 threads/core)16 Cores (8 threads/core)
8 Crypto Accelerators (1/core)16 Crypto Accelerators (1/core)
1 On Board PCIe x8 v1 Port2 On Board PCIe x8 v2 Ports

It is interesting to note that the T2 processor was only used in single socket systems. The T2+ processor removed the T2’s on board 10 GbE ports and other components to make room for the SMP glue. With the T3 processors, the 10 GbE ports have returned and the chip has built in glueless support for 4 way servers.

All in all they have packed more T-Series goodness in a smaller package but I’m not making goo-goo eyes yet.

For one, the smallest T3 based server, the T3-1, has the same number of threads as the T5140 but takes twice as many rack units. Although the T3-1 supports more PCIe cards and more internal hard disks, I would rather have a 1RU server or else have it support twice as much RAM.

The T3-2 server supports 256 threads. Compared to the T5440, it is actually smaller at 3RU and uses less power which sounds like a step in the right direction. Unfortunately, the T3-2 is also light on RAM supporting a maximum of 256GB compared to the T5440’s 512GB.

In short, The T3 series is a little off course for me at the moment. As a platform for consolidating tens of smaller applications, the thread to RAM ratio is too low making it hard to get 100% utilization out of these servers. With the T3-4 servers loading even more processing power into a single machine, the thread to machine ratio high as well. This is good if you are running a few really huge applications but if you are consolidating many smaller applications, you will not want to put this many eggs in one basket.

Sun Oracle Webcast Wrap Up

Last night I watched almost the entire 5 hour live webcast announcing Oracle’s strategies regarding the Sun Microsystems acquisition. As a near-evangelist for Sun and Solaris, I’m very happy with the deal finally going through and even happier that most of what Oracle said makes sense to me as a customer.

What I liked:

  • The clear commitment to the SPARC roadmap especially the T series. I honestly don’t know what I would have done if the T series servers disappeared. I’m very happy that they put raising the clock speed into the roadmap because some applications just can’t be deployed on these servers.
  • The clear commitment to making waves in Enterprise Storage. NetApp was specifically mentioned and obviously the 7000 series arrays are best suited to compete with the NetApp arrays but I hope they will draw some EMC blood as well. I like the plans for integrating backup capabilities.
  • The plans to integrate really great Solaris tech into Oracle applications like DTrace, and RBAC
  • The plans to offer direct support. Honestly this was one of the most annoying parts of working with Sun was having to work with different support providers in every location.
  • The plans to change the supply chain and ship direct- no more out of stock excuses.
  • The plans to integrate Ops Center with Oracle Enterprise Manager.
  • Larry Ellison’s stand up comedy
  • And completely unrelated- the flashing disk lights on the Exadata V2 🙂

I didn’t like:

  • The obvious cut planned for the x64 line of hardware. While they are keeping x64 where convenient (storage appliances, database machines, various other “clusters”) it looks like Oracle has no plans for dealing in x64 server business as a server business. I’m not a big user of the x64 stuff for servers but Sun doesn’t really offer anything reasonable for entry level anymore except the x64 line. This brings me to my next point-
  • The SPARC roadmap is slightly sucky as in how much processing power do you really want inside a single box.  According to the roadmap, their next plan is to double the amount of cores in a T3 processor so you’ll have one cpu with 16 cores and 128 threads. Their going to put two in a machine? four?  Here is how I see the servers they have today:
    • T1000- useless poorly designed server
    • T2000- ok server but a waste of rack space at 2RU
    • T5120- ok server but a waste of rack space considering I could put a T5140 in the same space
    • T5220- worse than the T5120 at 2RU
    • T5140- The best server ever built with exactly the right amount of everything
    • T5240- 2RU again???
    • T5440- I could serve ~8.64 billion web requests per day from one of these but I’d need a 1.6Gbit uplink and two servers for redundancy = 8RU, or else use 4 T5140 machines, deliver the same performance, and use 4RU?- maybe 5RU including n+1 redundancy.
    • NONEXISTANT – little SPARC machine for backup/monitoring/insert your SPARC only app that doesn’t deserve a minimum of 32 threads and 2RU  here.

    At some point, you just want more smaller machines for less points of failure. I really have uses for low end SPARC machines and they don’t make them any more.

  • I don’t really like the “server phone home” idea.
  • No mention of OpenSolaris- I’m not really a user but I didn’t like that it wasn’t mentioned- What does that mean??
  • No mention of Webstack. I really like Sun Webstack as an idea. I’m not sure what is happening to it now?
  • No mention of how Oracle will be combining the knowledge bases? Sunsolve? Bigadmin? docs.sun.com? forums.sun.com (looks like this already had an Oracle makeover :?)

One thing I’m not sure about is the integration of Sun virtualization technologies into Oracle VM. On one hand it sounds good, on the other hand, I think this was the only part of the presentation where I noticed there were no due dates. Virtualization is super important to me so I really want to know where things stand.

Obviously, it is easy to  get up and say everything will integrate but doing it is much harder. Just getting past the internal politics of this will be a major issue. Now we can only wait and see if Oracle can pull it off.

I used to get upset with “Oracle people” for always thinking that Oracle was the solution to every problem. If they pull off this acquisition, I much just become an “Oracle person” myself.

EMC Replication Manager in Solaris

UPDATE: No ZFS Support for Replication Manager in the near future

Using storage level snapshots can be used to run backups without directly requiring resources from the original host.

EMC Replication Manager coordinates the creation of application consistent snapshots across all the hosts in your network. It handles scheduling creation/expiration of snapshots,  mounting and unmounting from backup servers, etc. from a single console.

Although it is not tightly integrated into EMC Networker like the similar Networker PowerSnap module, it can be used to start a backup process after taking a new snapshot and it has the capability to manage snapshots unrelated to backups from a GUI.

While the data sheet claims support for Solaris, there are several caveats which I have run into.

  1. There is no mention of ZFS support in the data sheet and apparently, there is no support in the software either. One would expect this to be a non-question since ZFS has been part of Solaris since 2006.
  2. The data sheet is missing the word “SPARC” next to the word Solaris. There is no support for x86.

Honestly, this has put a dent in my plans since my backup server is an x86 box. I’m hoping the lack of ZFS support will work out as long as we can script any FS specific magic we need. I don’t have an option of running something like Linux on it (just to get the software working) because I won’t be able to even mount the ZFS filesystems- let alone back them up.

In the meantime, I’ll have to move my backups to a SPARC server and considering the lack of low end SPARC machines, I’ll have to allocate something way too expensive to be a backup server.

Sun’s Predicament

I’ve been working with Unix for a fairly long time now- about 13 years.

I’ll admit that I started with Linux and thought it was light years ahead of SunOS 4.x running on those old SPARC machines- I mean who had heard of SPARC processors? I remember my boss trying to explain to me that even an older SPARC processor was more powerful than a newer Intel Pentium processor. I didn’t really believe him. In time, I convinced them to get rid of most of their SPARC/Solaris in favor of the hip, free, and cheap Intel/Linux combination.

Now I see that I couldn’t have been more wrong. I realize that SunOS 4.x probably still has features which I don’t know how to use properly. When I look at Solaris 10, ZFS, Zones, LDOMS, DTrace, etc. I not really sure you could pay me to work with Linux (that would be soo depressing). That isn’t even mentioning the SPARC hardware it runs on- Can any Intel server compare to a T5140???

That’s why the current situation with Sun absolutely SUCKS (pardon my french)! I’m sure there are a lot of admins out there who feel the same way. If this Oracle deal doesn’t go through and Sun disappears because of it, it will be our loss. We’ll be stuck with mediocre operating systems and commodity hardware and I really hope it doesn’t happen.

That said, I’d like to say thanks to all the people at Sun who are still turning out crazy cool technologies despite the problems.

Listing ZFS Clones using the origin property

Recently I created my first ZFS clones but quickly realized that there was no simple way to tell the clones from the regular filesystems. My first instinct was to run ‘zfs list -t clone’ similar to ‘zfs list -t snapshot’ but this didn’t work. Maybe it works in newer versions of ZFS.

After some poking around I found the ‘origin’ property which sets the clones apart so running something like-

zfs list -o origin,name,used,avail,refer,mountpoint | \
grep -v ^- |awk '{print $2"\t"$3"\t"$4"\t"$5}'

will get you what you are looking for.

If you haven’t played with ZFS clones yet, basically they are writable snapshots of a file system.

They are great if you want to copy a lot of data to the side, modify it, and possibly replace the original data, without taking a lot of time or disk space. The ZFS clones take seconds to create, since they don’t actually copy any data, and they will only store the blocks which have changed since their creation. If you want to replace the original data, you can then transparently promote the clone to be the master filesystem and turn the master into a clone.

The downside of clones is that they are always dependant on the snapshot from which they were created. You can not destroy a snapshot on which a clone is based without destroying the clone.

For the sake of simplicity and since I don’t usually have disk space issues, I usually prefer to make full copies using ZFS send/recieve but I have definate plans to make more use of ZFS clones in the future.