pureScale on Linux
A quick word on what circumstances pureScale is best suited to.
First to say what it is not not suited to i.e. data warehouse type applications. It is a shared disk solution and as such not really suitable for data warehousing. This is because of tendency of large transactions being the main workload in such an environment.
It is suited to OLTP loads.
Do you need to come up with a database solution for your application? This could be a new build or replacing old hardware and software.
Do you have an application that generates a lot of small of smallish transactions?
Do you need continuous availability and built in resilience?
Do you need to be able to ramp up the capacity of your system easily in the future rather than buying all of the hardware and licenses you might need over the next 2 - 5 years now?
If the answer if yes to most of these questions then pureScale is for you.
Marco Bonezzi 270005PS6A 2,656 Views
Last week during some performance tests on one of our PureData for Transactions Systems, we were looking at the CPU load for the system during workload execution.
In pureScale you know that the CPU stats from command line tools like top or vmstat could be not accurate to the DB2 member real load. This is due to the Cache Facility (CF) process being colocated to the DB2 member, on the same host. The process for the CF will shows most of the CPU usage even if there is not to worry about this.
To clarify this, we were using a small instance on PureData for Transactions, so 2 members and 2 CFs on two physical hosts.
Back to our CPU load... How can we measure that?
db2top? There are some CPU metrics in db2top, but can show 100% CPU load even if this is not exactly true.
Optim Performance Manager? Yes, we can check the CPU load for the system using OPM, in the Performance - Overview - Key indicators panel. We can also look at Performance - System to get the CPU usage. In PureData for Transactions this is useful as it has Optim Performance Manager already integrated into the management console.
An useful query to check the CPU load from DB2 itself and also valid for pureScale is the following:
select MEMBER, varchar(HOST_NAME,32) as HOST_NAME,
[db2sdin1@compute01 ~]$ db2 "select MEMBER, varchar(HOST_NAME,32) as HOST_NAME,
Other useful information like this can be also gathered using SYSPROC.ENV_GET_SYSTEM_RESOURCES function:
NEW Information Management Solution Portal!
The new site features:
We have heard that a “tens of kilometres” limit applies to the distance between the two sides of a Geographically Dispersed pureScale Cluster (GDPC). Buy why?
This is based on a physical limitation i.e. the speed of light in glass (fibre) which is about 5 µs / km. From this we can calculate a round-trip times from member to CF as follows at these distances:
3km = 30 µs
10km = 100 µs
50 km = 500 µs
100 km = 100 µs (or 1 ms)
300km = 3000 µs (or 3 ms)
This will have a significant effect on the performance of the cluster, especially when we start to get into tens of kilometres. The “normal” times for RDMA actions are of the order of 15 µs so to those we need to add this latency for the distance. Compared to a normal pureScale cluster (all in one location) an RDMA action will be slower at a distance as follows.
3km = 3 times slower
10k = 8 times slower
50k = 33 times slower
100k = 66 times slower
300k = 200 times slower
µs = microseconds or 10-6 seconds
ms = milliseconds or 10-3 seconds
CiaranDeB 2700033FRG 2,364 Views
I am at the pureSystems booth at the Innovate 2013 conference in Orlando. http://www-01.ibm.com/software/rational/innovate/ This is my first big conference and I have to say I am very impressed by the scale of things! Below a rendering of our setup and we are just a tiny part of this conference!
We have three touchscreen monitors loaded with the TouchScope application. These are pretty cool. We have experts in each of the types of pureSystems at the booth also who can give a great insight into the different types of pureSystems: i.e. pureFlex, PureApp and pureData.
CiaranDeB 2700033FRG 2,201 Views
CiaranDeB 2700033FRG 2,136 Views
The more I talk to people about pureSystems the more I think that the old way of provisioning systems is nuts. Do people really need to design a solution, buy various "bits" like storage, networking, servers seperately from different vendors with different support agreements, cable it all up and put a stack of software on there "manually" themselves?
If you went to buy a car and they said you had to do this: figure out what parts you need, then buy the engine from one company, the wheels from another, the chassis from yet another etc and then put it together yourself and hope it all works ok you would think they were crazy, right?
pureSystems the smarter way to buy IT! http://www.ibm.com/ibm/puresystems/us/en/index.html
CiaranDeB 2700033FRG 1,933 Views
Outline spec of what we are currently working on...
Make the cluster setup as "production like" as possible and to be able to walk clients through the various features of this.
Be able to easily adjust the workload profile to have different mixes of short read, update, insert workload and longer ad-hoc queries to be similar to a clients's real workload.
Base the demo workload on a well known benchmark for OLTP systems (TPCC).
Using the Technology explorer front end.
Demonstrate workload balancing scenarios:
Client affinity for ad-hoc queries
Mixture of the above
Demonstrate failover scenarios and demonstrate continuous availability:
Member fail, various reasons
Standby CF fail, various reasons
Multiple failures, various reasons
Reproduce failover scenario and demonstrate minimal downtime:
Main CF fails
Monitoring information on Db2 on members , Show graphically second by second:
OPM dashboard is a useful overview on an aggregate (minute by minute) basis.
Workload/throughput information. Show graphically second by second:
Transactions per member for read, Non-read (insert, update, delete)
Overall throughput of transactions.
Automate the setup and tear down as far as possible.
Document the setup and running of the demo so that we can reproduce and also so that others can.