Chat Transcript: April 2016 Workload Automation Autosys Edition (AE) Office Hours

Document created by Lenn Thompson Employee on Apr 18, 2016Last modified by Lenn Thompson Employee on Apr 20, 2016
Version 2Show Document
  • View in full screen mode

Lenn Thompson (CA): Good morning, everyone and thanks for joining us. What questions do you have for the AE team?

pavel: how is weather in Plano?

Jennifer Barber: GM Lenn

Jennifer Barber: I saw on the forum this morning that Oracle 12c container was not supported. I haven't looked into this yet. Does this mean that RAC is not supported?

Lenn Thompson (CA): @Jennifer: Do you have any questions for the team?

Lee Peterson (CA) (CA): @Pavel - I'm assuming its good.I'm near Plano IL, and its about 50 degrees

Lee Peterson (CA) (CA): @Jennifer- no, RAC is different

Jennifer Barber: ok, good - just checking

Lee Peterson (CA) (CA): @Jennifer - sure, it was a fair question

Jennifer Barber: I just saw this meeting this morning, so I didn't come prepared with any questions - lol

Lee Peterson (CA) (CA): @Jennifer - yea, me too

Jennifer Barber: I think it's a great idea though

Lenn Thompson (CA): @Jennifer: Well thanks for joining us. Maybe something else will spark an idea!

Jennifer Barber: yes

Lenn Thompson (CA): Are there any other questions?

Jennifer Barber: can you talk a little about what reporting and critical path monitoring capabilities come out of the box with WCC 11.4?

pavel: i wonder about Ideas review process.

Jennifer Barber: or is that too much for this forum

pavel: how often the review takes place?

pavel: there are many ideas in with status "New", yet they are more than one year old.

Dan Shannon (CA): @Pavel Ideation reviews haven't taken place as frequently as intended. The goal is to review them monthly. We're in the process of aggressively addressing this. You can expect to see numerous changes in the next few weeks

pavel: ok

Lee Peterson (CA) (CA): @Jennifer -The reporting you are speaking of is more of a topic for iDash. The iDash analytics will probably be what you are looking for. WCC is a little light in terms of the analytics you are inquiring about - it’s not really as depthful as iDash.

Jennifer Barber: ok thanks

pavel: What release will support kerberos authentication with Event Server db (sybase in my case)?

Jennifer Barber: we are looking at iDash too, but I thought I had heard there was some capability out of the box

Lee Peterson (CA): @Jennifer - There is not any mechanism to monitor the critical path like iDash, outside of your creating monitoring views based on your own understanding of critical path.

Jennifer Barber: ok, thanks

Lee Peterson (CA): @Jennifer - sure thing

Jennifer Barber: I should be installing it any day now anyway

Lee Peterson (CA): @Pavel – it’s not currently planned

pavel: i though otherwise. thanks

Lee Peterson (CA): @Pavel - deferring to Dan

Dan Shannon (CA): @Pavel I'm not sure what expectations were set. The plan is to meet all previously agreed promised decisions

pavel: @Dan, I have no expectation, just wonder when this idea might be delivered.

pavel: https://communities.ca.com/ideas/107294584

Lenn Thompson (CA): Keep the questions coming, everyone. Thanks for joining us.

Lenn Thompson (CA): I also want to draw your attention to our CA World '16 Call for Speakers. We'd love you to share your experience with AE. https://communities.ca.com/community/ca-world/blog/2016/04/08/ca-world-16-call-for-speakers-is-now-open

pavel: I’m looking for a "not sooner than" answer. for example not sooner than 2017

Lee Peterson (CA): @Pavel - please stand by, we're reviewing your link

Dan Shannon (CA): @Pavel I see the status is currently planned. I'll look into the prioritization of this and get back to you offline

pavel: ok. what release will come after r11.3.6 SP4?

Dan Shannon (CA): @Pavel We currently have plans for 11.3.6 SP5 & SP6

pavel: thanks

pavel: Next question re: EnableIPCaching=0

pavel: how much overhead/delays it might cause?

Lee Peterson (CA): @Pavel - lay it on me

Lee Peterson (CA): Delays?

Lee Peterson (CA): @Pavel - it’s actually really beneficial in many situations. There is a lookup table kept in memory, which negates the need to even make an getaddr system call. So the gains are quite substantial, even over local nscd.

pavel: I assume IP address caching is enabled by default to speed thing up..

Lee Peterson (CA): @Pavel - that being said, there can be some situations, where ip's are dynamic, and that creates a problem. Caching is on by default.

pavel: understand the benefit, but trying to understand negative impact of EnableIPCaching=0

karen: Hello all, what are the plans for r12?

Dan Shannon (CA): @Karen We're currently working on delivering features quicker on top of 11.3.6 via service packs opposed to waiting for a new release. We are still planning to produce r12 but it has been deferred in favor of the SPs right now

Lee Peterson (CA): @Pavel - the negative is going to be when you have a dynamic ip.If our table in memory is stale, you may have to receive an error before we update the table. Do this enough times and you have some poor performance. So I typically see problems with ipcaching in environments where they heavily use dynamic ip.

Lee Peterson (CA): @Pavel- btw, in those situations where EnableIPCaching is bad - use something like nscd. That’s a very happy medium as it gives you the autonomy of refreshing or flushing the name cache, independent of Autosys.

karen: @Dan, thanks!

pavel: interesting. so, nscd can compensate for disabled IP caching?

Lee Peterson (CA): @Pavel - quite well, yes. However, autosys is still having to make the getaddr system call, which is still some overhead for name resolution.

Lee Peterson (CA): @Pavel - so if you can use the AutoSys ip caching, that is the fastest.If you cannot, nscd may be the next best thing.

Dave: Questions regarding Box Job usage

pavel: i think extra 0.02s for name resolution is not bad.

Dan Shannon (CA): @Dave what would you like to know about boxes?

Dave: I'm already aware of the following Best Practices

Dave: 1. No start times on jobs in Boxes

Lee Peterson (CA): @Pavel - these things are relative and can compound, but I can appreciate the freedom that nscd would give you.

Dave: 2. No dependences on jobs in Boxes to jobs in other Boxes

pavel: thanks

Dave: User's Guide talks about using Boxes for similar starting conditions - not organizationally

Dave: what potential problems using Boxes organizationally?

Dave: in other words, Best Practices related to using Boxes organizationally

Lee Peterson (CA): @Dave - just to add. Organizationally being the starting conditions. Application and Group are other options for "organization" from a logical sense. You're spot on though.

Jennifer Barber: this is one of my personal pet peeves, Dave - lol... I'm curious to see CA's response

Lee Peterson (CA): @Dave- The problem with "abusing" boxes is that they will cause substantial overhead and race condition risks. The intended design is to keep things based on common dependencies that can be set at the parent level. When excessive nesting and redundant conditions shared by both child and parent are used, you'll run into trouble.

Lee Peterson (CA): @Dave - diminishing returns on boxes typically begins at 4 deep. Beyond 5 will start to create problems.

Dave: Hi Lee

Dave: I'm aware of the "No nested boxes" rule

Dave: If there are no nested boxes, are there any other problems with using Boxes organizationally?

Lee Peterson (CA): @Dave - sometimes it’s unavoidable, but there is a point where it’s excessive. Problems, well not really, but we offer "Application" and "Group" attributes for organizational requirements. Those are probably more appropriate. At least, from a CA perspective - this is what we have in mind when designing the product.

Dave: Thx Lee

Lee Peterson (CA): @Dave - you're welcome!

Dave: Guess I'm looking for something other than the philosophical response

Jennifer Barber: in my experience, it works well if there are similar starting conditions

Dave: in other words, using boxes for starting conditions and not organizationally will result in avoiding problem x,y,z

Jennifer Barber: but if you have jobs waiting for different run windows, different start times, or dependencies outside the box, you can run into trouble

Lee Peterson (CA): @Dave - well, was mine too philosophical? The risks are really in the use of conditions and nesting. Otherwise, it’s not unusual for some customers to use boxes for organizational requirements, but there are attributes available which are there for that very purpose.

Jennifer Barber: agreed, Lee - using application or group attributes work well for organization

Dave: "The risks are really in the use of conditions" --> Could I trouble you for an example?

Lee Peterson (CA): @Dave - sure - let me write you a jil example

Mark Hanson: Jeopardy theme

Jennifer Barber: lol

Lee Peterson (CA): @Dave:

insert_job: box

job_type: box

date_conditions: 1

days_of_week: all

start_mins: 5,10,15,20,25,30,35,40,45,50,55,0

insert_job: command1

box_name: box

machine: host.company.com

command: echo "hello Dave"

date_conditions: 1

days_of_week: all

start_mins: 5,10,15,20,25,30,35,40,45,50,55,0

insert_job: command2

box_name: box

machine: host.company.com

command: echo "hi Lee"

date_conditions: 1

days_of_week: all

start_mins: 5,10,15,20,25,30,35,40,45,50,55,0

insert_job: box2

job_type: box

box_name: box

date_conditions: 1

days_of_week: all

start_mins: 5,10,15,20,25,30,35,40,45,50,55,0

insert_job: command3

box_name: box

machine: host.company.com

command: echo "hi Lee"

date_conditions: 1

days_of_week: all

start_mins: 5,10,15,20,25,30,35,40,45,50,55,0

Lee Peterson (CA): ^ An example - you can add condition to reference internally, or externally. The medium here (webex) is probably not the best. I can offer more detail and better formatting if there's anyone here who wishes to receive an offline response on the topic.

Lenn Thompson (CA): Okay everyone, we only have a couple minutes left and I some of the folks on the AE team have a meeting starting at the top of the hour. Are there any other last-minute questions?

Greg Wade: I don't want to derail this discussion. Are all parts of WA AE compatibile with Win 2012R2 64bit?

Lee Peterson (CA): @Greg- CSAM/EEM/WCC/AutoSys/EM/Agent/Remote - let me get you the link to our matrix.

Jennifer Barber: I look forward to you guys doing this again - I really like this format

Greg Wade: EM was my last holdout.

Thompson (CA): @Jennifer Thank you for that feedback. Thanks for joining us!

Jennifer Barber:

Lee Peterson (CA): @Greg, EM is probably going to be the least likely. Send me an emailpetle02@ca.com. We're running out of time.

Dan Shannon (CA): @Greg EM does not run on Windows 2012 R2

Greg Wade: Thanks.

Lenn Thompson (CA): We need to wrap, everyone. Thanks again for joining us. Have a great day. We'll see you next time and in the community!

karen: Thanks!

Attachments

    Outcomes