c0t0d0s0.org

When you size storage, you have to size for three things: IOPS, capacity and bandwidth. But that has an interesting implication. When we had 2,4 or 9 GB harddisks we needed many disks to provide the needed capacity and thus had often enough IOPS for the tasks of that time without needing further provisions. Then disks got bigger, and we created massive storage arrays with much to large capacity , but we did it for the IOPS. Now we have SSD in the TB range and with IOPS north of 100.000 and you would think that your problems are gone, however now the third parameter comes into the game: Bandwith.

So far you had often enough disks (either because of bandwidth or IOPS requirements) and thus enough raw bandwidth at least at the point right behind the disk). However it doesn't matter, when you have 1 TB and 100.000 IOPS, you still have just a storage connection that limits your transmission at 6 GBps, come hell or high water. So after explaining "No, a 1 TB USB disk from el-cheapo-component-shack isn't enough for the 1 TB database, because of IOPS yadda yadda yadda" to the people in the purchasing department, we probably have to start to explain "No, the 1 TB SSDs aren't enough as well, because of the bandwidth requirements yadda yadda yadda"

Seemingly simple topics are never simple. Just collecting some stuff about the topic "Swap. How to size it". Well ... it isn't simple ... not at all. Especially when you want to explain, not just giving a rule of thumb ...

On April 25th an webbased event about Solaris 11 takes place: It's named Oracle Solaris 11: What’s New Since the Launch.

Agenda
9:00 a.m. PDT	Keynote: Oracle Solaris - Strategy and Update Markus Flierl, Vice President, Oracle Solaris Engineering
9:40 a.m. PDT	Oracle Solaris 11: Extreme Engineering - A Technical Update Dan Price, Senior Principal Product Engineer, Oracle Solaris Engineering Bart Smaalders, Senior Principal Product Engineer, Oracle Solaris Engineering
10:20 a.m. PDT	Customers and Partners: Why We Moved to Oracle Solaris 11 A discussion of the reasons why businesses and commercial software developers have adopted Oracle Solaris 11, from the people responsible for these decisions
11:00 a.m. PDT	Oracle Solaris: Core to the Oracle Systems Strategy John Fowler, Executive Vice President of Systems, Oracle

9:00 am PDT is 18:00 in Berlin, 17:00 in London and i assume much to late in Tokyo with 01:00 am the next day ...

"How to Update Oracle Solaris 11 Systems Using Support Repository Updates" is a great article written by Glynn Foster in order to explain the use of the Support Repository for Solaris 11. A must read.

Am 26. April findet das zweite Oracle Oracle Breakfast in Hamburg in der Geschäftsstelle (Kühnehöfe 5) statt: Also Futtern mit technischem Content. Auch diesmal gibt es zwei Vorträge.

Agenda
9:30	Willkommen zum Frühstück
10:00	Solaris 11 im Detail - Einbindung in heterogene Netze (CIFS-Dienst etc.) Joerg Moellenkamp
11:30	Kaffeepause
12:00	ZFSSA praktisch Einbindung einer ZFS SA in heterogene Netze, aber wie? Vortrag & Livedemo unter VirtualBox Dirk Nitschke
13:30	geplantes Ende

Anmelden könnt Ihr euch mit einer formlosen Mail an oraclebreakfast_ham@c0t0d0s0.org. Das ist ein Forwarder an die Addresse der Kollegin, die das intern bei uns organisiert, deren Mailaddresse ich nicht unbedingt für Spammer verteilen möchte ...

After having some discussions i now made my mind about it: In the next weeks you will see many republications of old articles in the blog as i will republish all articles in the LKSF, however checked and updated for Solaris 11 (some Opensolaris based stuff in the lksf is working slightly different, and if it's just for different package names). However this will take time, as i will do this on weekends and evenings. At the end i will just recollect them and create a Solaris LKSF pdf again.

Ruud van der Pas and Jared Smolens wrote an really interesting whitepaper about the SPARC T4 and its behaviour in regard with certain code: How the SPARC T4 Processor Optimizes Throughput Capacity: A Case Study. In this article the authors compare and explain the behaviour of the the UltraSPARC T4 and T2+ processor in order to highlight some of the strengths of the SPARC T-series processors in general and the T4 in particular.

I was really astonished when i saw this question. Because this question was a old acquaintance from years ago, that i didn't heard for a long time. However there was it again. The question: "What's the overhead of Zones?". Sun was and Oracle is not saying "zero". We saying saying minimal. However during all the performance analysis gigs on customer systems i made since the introduction of Zones i failed to measure any overhead caused by zones. When there was overhead in that situation it was in the error of measurement.

What i saw however, was additional load intoduced by processes that wouldn't be there when you would use only one zone Like additional monitoring daemons, like additional daemons having a controlling or supervising job for the application that resulted in slighly longer runtimes of processes, because such additional daemons wanted some cycles on the CPU as well. Like additional sshd daemons, like additional init processes. So i ask when someone wants to tell me that he measured a slight slowdown, if he or she has really measured the impact of the virtualization layer or of a side effect described above.

It seems to be a little bit hard to believe, that a virtualisation technology has no overhead, however keep in mind that there is no hypervisor and just one kernel running that looks and behaves like many operating system instances to apps and users. While this imposes some limits to the technology (because there is just one kernel running you can't have zones with different kernels versions running ... obvious even to the cursory observer), but that is key to it's lightweightness and thus to the minimal overhead.

Another Unix on a very small device: "UNIX ON PIC32 – MEET RETROBSD FOR DUINOMITE". The Duinomite is a small board based on a PIC32 microcontroller.

Serge did amazing job by porting the old days 2.11BSD Unix used to run on PDP-11 to PIC32 (MIPS). In just 128KB RAM footprint he manage to boot UNIX OS and you have 96KB left for applications.

It is really easy to configure an NTP client on a plain standard Solaris 11 (that said it isn't more difficult on Solaris 10 ... it's the same procedure). So i will just give a very short walkthrough.

As you may have noticed many configuration tasks around name services have moved into the SMF in Solaris 11. However you don't have to use the svccfg command in order to configure them, you could still use the old files. However you can't just edit them, you have to import the data into the SMF repository. There are many reasons for this need but the ultimate one is in the start method. I will explain that later. In this article i want to explain, how nscfg can help you with with the naming service configuration of your system.

Alan Coopersmith gives some interesting and important insights into the future of Solaris.