Load balancing in Autosys


 
Thread Tools Search this Thread
Top Forums UNIX for Advanced & Expert Users Load balancing in Autosys
# 1  
Old 07-14-2014
Load balancing in Autosys

Hi,

I am working on development project where I have to migrate many jobs from Tidal to Autosys R11.

During this project we came across the following requirements.
1. There are 3 real machines. There could be many jobs activated simultaneously, but only one job should execute at a time and it should select any machine out of 3 machines.

2. There are 3 real machines. There could be many jobs activated simultaneously, but only 3 jobs should execute at a time. The jobs should be equally distribute on all machines.

We have used "job_load" and "max_load" features in Autosys R11 and managed the solution for 2nd case. But 1st case is unsolvable.
Also we faced issue when only 1 job running at a time because it is always choosing first machine to run.

We have used virtual machine to club real machines.

We are also evaluating options in R11.3. There we are planning to use "resource" attribute.

Still if anyone has faced such scenarios and if anyone has solutions on these points then please help us.
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. Shell Programming and Scripting

Load Balancing in UNIX

Dear All, Can any one help me for this request? There is a case. I have 20 files which I need to FTP to 5 servers. I want to know if there is any possibility to make a load balancer which transfers files in round robin manner to 5 servers. As per theoretical algorithm, what I think, flow can... (9 Replies)
Discussion started by: Zaib
9 Replies

2. UNIX for Advanced & Expert Users

Help in MQ load balancing

Hi, Currently we have 3 old and 3 new servers catering to Live traffic. As my component move from legacy interfaces to MQ one, we want to have load balancing of old interfaces available on MQ interface as well. For this, we want to send only 30% of all MQ traffic on 3 OLD Live servers, and want... (1 Reply)
Discussion started by: senkerth
1 Replies

3. IP Networking

Load Balancing ppp

Hello everybody How can i Load Balance two slow ppp(gprs) connections with iptables . (4 Replies)
Discussion started by: rink
4 Replies

4. Linux

HTTP load balancing.

Hi, We have 2 pools of servers. Lets call them A and B and they would contain 2 servers each. Pool A will be hosting www.example.com/app/v1 and pool B will be hosting www.example.com/app/v2. Clients will be requesting right url (/v1 or /v2) but will be hitting just one IP. I'd like to: 1)... (3 Replies)
Discussion started by: chrisfb
3 Replies

5. UNIX for Advanced & Expert Users

High availability/Load balancing

Hi folks, (Sorry I don't know what its technology is termed exactly. High Availability OR load balancing) What I'm going to explore is as follows:- For example, on Physical Servers; Server-1 - LAMP, a working server Server-2 - LAMP, for redundancy While Server-1 is working all... (3 Replies)
Discussion started by: satimis
3 Replies

6. Web Development

Load Balancing in Apache

Hi All, I have one webserver which has an application for a set of internal users can be accessed by _http://server1.com I am planning to load balance this application. For that I have cloned this server and build a new one which can be accessed using _http://server2.com]Server2.com. Also i... (2 Replies)
Discussion started by: Tuxidow
2 Replies

7. Solaris

Load balancing with IPMP

Is it possible to do a load balancing ( incoming and outgoing )with with IPMP in solaris 10 like sun trunking ? If yes what are the steps involved in it , i know how to do the failover IPMP both link based and probe based but i 'm looking for possible load balancing (3 Replies)
Discussion started by: fugitive
3 Replies

8. Ubuntu

perlbal and load balancing

Hi guys, I wonder if someone would be able to assist with my problem. I have just set up a load balancer for a company I am working for. HTTP redirection is working fine, however they also want to load balance SSH and FTP too. At the moment the perlbal config looks like; CREATE POOL webhttp ... (1 Reply)
Discussion started by: JayC89
1 Replies

9. AIX

idea on hacmp load balancing

Hi All, Any idea about load balancing on hacmp? Or load balancing is only on lpar. Any idea or link info will do. Thanks in advance. (2 Replies)
Discussion started by: itik
2 Replies

10. UNIX for Dummies Questions & Answers

Question about load balancing

If you have two or more servers load balancing, are the servers mirroring one another? If images, etc., are uploaded, will they be stored on all the servers so that if one server goes down, the images will be served up by another server? (1 Reply)
Discussion started by: wvmlt
1 Replies
Login or Register to Ask a Question
job(n)								 Tnm Tcl Extension							    job(n)

__________________________________________________________________________________________________________________________________________________

NAME
job - Invoke Tcl procedures at regular intervals. _________________________________________________________________ DESCRIPTION
The job command provides a convenient mechanism to implement Tcl procedures that are invoked at regular intervals. Jobs are activated from the Tcl event loop. Every job can have arbitrary attributes to store job specific state information between two invocations. These attributes help to avoid global variables in order to keep the Tcl name-space clean. The job scheduler itself relies heavily on the Tcl timer interface which itself depends on the system clock. Moving the system clock back- wards can have the effect that jobs are not activated for the amount of time the system clock was moved backwards. Similarly, moving the system clock forward can have the effect that jobs are activated earlier for the amount of time the system clock was moved forward. JOB COMMAND
The job command allows to create new jobs and to wait for jobs to complete. The job command can also be used to retrieve status information from the job scheduler. job create [option value ...] The job create command creates a new job object. The options can be used to configure the behaviour of the new job. See the config- ure command below for the details. job current The job current command returns the name of the currently running job. An empty string is returned if no job is running. job info The job info command returns a list of all job names. The list will be empty if there are no jobs known in this Tcl interpreter. job schedule The job schedule command invoke the scheduler. This is usually done automatically from the event loop. This command may be used dur- ing long computations since the scheduler is not preemptive. Note, this command can cause arbitrary side effects and should be used with care. job wait The job wait command blocks until all existing jobs have been finished. The command will wait infinitely if there is at least one job left. Events are processed while waiting for the jobs to finish which can have arbitrary side effects. JOB INSTANCE COMMAND
Every job is represented by a job object command which allows to manipulate the job. job# attribute [name [value]] The job# attribute command allows to save job specific data in job attributes. If called without any arguments, all existing attribute names for this job will be returned. If called with a name argument, the current value of the attribute will be returned. Non-existing attributes do not produce an error. Instead, an empty string is returned. Calling the option with a name and a value causes the value to be saved in the job attribute. job# cget option The job# cget command returns the current value of a configuration option. See the description of supported options below. job# configure [option value ...] The job# configure command manipulates configuration options. See the description of supported options below. A list describing all current options is returns if the options argument is missing. job# destroy The job# destroy command destroys the job object. This command forces the object to change into the expired state. The scheduler will cleanup this object the next time it is active. job# wait The job# wait command blocks and processes events until the current job changes into the expired state. JOB OPTIONS
Every job has associated configuration options which control when a job is activated and which Tcl command is bound to the job. -command command The -command option defines the Tcl command that is evaluated whenever the job is activated. -interval time The -interval option defines the time interval between two job activations in milliseconds. -iterations number The -iterations option defines the total number of times that a job is activated. If this value reaches 0, the job will change its state to expired. Note, if the iterations options is never used, it will also have the value 0 but nothing special happens to the job object. -status state The -status option provides access to the current job state. A job is always in one of the states waiting, suspended, running and expired. A job in the expired state will be removed from the system once the scheduler gets activated again. Suspended jobs are not activated by the job scheduler. A suspended job can be resumed by changing its state to running. -time time The -time option returns the number of milliseconds until this job is activated the next time. This option is read-only. SEE ALSO
scotty(1), Tnm(n), Tcl(n) AUTHORS
Juergen Schoenwaelder <schoenw@cs.utwente.nl> Tnm job(n)