CA Performance Management Office Hours Transcript - September 2015

Document created by Melanie_Giuliani Employee on Sep 22, 2015Last modified by SamCreek on Dec 17, 2016
Version 2Show Document
  • View in full screen mode

Chat transcript from CA Performance Management Office Hours: A Live Online Chat [September 2015] which took place on September 22, 2015.


Melanie Giuliani to Everyone: Hi everyone - welcome and thank you for joining! Let's get started! Please type your questions into the chat box.

Melanie Giuliani to Everyone: We have CA Support standing by. Does anyone have any technical questions?

Daniel Holmes to Everyone: While you guys are thinking of your questions - i will share an update on CA PM 2.6 - we are in our last week of Early Availability and if all goes well we are excited to GA

Alexander Bradley to Everyone: Do Data Collectors write to the Vertica database directly?  I know Data Aggregator is involved in the transaction, but I don't understnad whether it is more of a faciliator or a go-between between the Data Collectors and the Data Repository

Michael Poller to Everyone: @Alexander - Data is passed from the DC, to the DA for insertion into the Vertica DR DB

Alexander Bradley to Everyone: ok, thanks @Michael

Michael Poller to Everyone: @Alexander - It’s a bit of both. The DA provides ties to CAPM UI for config work and changes, and to provide data back from Vertica from CAPM report requests for data. It also ties to the DC to tell it what to poll per discover and cert changes via CAPM UI. And then it as said before also takes polled data from DC to insert into DB

Catalin Farcasanu to Everyone: I want to know if there are any plans to update the available list of presented units in the interface. I have requests to display data that extends beyond existing units.

Mike Collins to Everyone: Hi, we are trying to get business hours working with normal (not site) groups, by adding the business hour into the dashboard and this is working when we do charts (seeing grey zones corresponding to the BH) but when doing a table it always says no data , even choosing 30 days time period ... on top of that there are no calculations made by the dashboard to actually calculate the metrics for BH periods. All works fine when we do a site group but then all components fitting the regex definition get tagged with that BH and this in any dashboard it is contained

John Murdough to Everyone: @Catalin - could you give an example of what you would like to have for additional units?

Catalin Farcasanu to Everyone: I would also be very interested in having some advanced examples in the Power Certification User Guide that should describe usage of MVEL for evaluation of data extracted from systems.

Mark to Everyone: When is the next release scheuled for CA PC? I'm looking for the fix for the problem with the NFA chart not loading correctly on the initial page load in the I.E 11 broswer.

Daniel Holmes to Everyone: @Mark the next release CAPM 2.6 is now in Early Availability and if all goes well this week we will GA

Catalin Farcasanu to Everyone: @John: Sure, I can quote you from the requests I received so far: kWh for example. Another one would be amperes, watts, all sort of power monitoring related units.

Mark to Everyone: @Daniel, thank you. 

John Murdough to Everyone: @Catalin - Relative to power cert guide.  Thanks for the input I'll forward that request to our docs team to add some content or reference material

Daniel Holmes to Everyone: @Mike Collins - right now business hours is designed to work with site groups as the geography of the physical location of devices typically dictates the timezone of the infrastructure and this is set on site groups - can you provide an example where you woudl like to set business hours on Generic Group type? We will take this under consideration.

Sean Moriarty to Everyone: @Mike - Business hours are built around the premise that the device has been given a site group association, such that the system understand that devices is within a given time zone and has a business hour filter. Without that association the system can not determine what device has been assoicated to a time zone profile.

Sean Moriarty to Everyone: Now we do allow the device to be within a time zone but not a business hour via a site group and that allows for a dynamic association to be applied but still need that device to be mapped to a site group.

Catalin Farcasanu to Everyone: @John: thanks for that. If you're trying to certify some devices that present data as Strings or OctetStrings, it's a pain to ge it working with the existing examples. It's such an uncommon situation.

John to Everyone: Question in regard to 2.6... Will there be more integration with Spectrum; in regards to administration with the default domain? Right now items are added to DD in Spectrum and populate in PC; but a deletion of a device/remove from DD requires a manual deletion in PC...

Catalin Farcasanu to Everyone: @John: *not such an uncommon situation.

John Murdough to Everyone: @Catalin - yes we've encountered that issue before with some metrics, I'm reaching out to see if we have some good reference examples

Melissa Potvin to Everyone: hi everyone! if you are new to office hours and/or just joining  -- welcome! we have support, engineering and product management on the line waiting to answer your questions -   use the chat window to submit questions.

Daniel Holmes to Everyone: @John right now we are targeting CAPM 3.0 (Spirng 2016) to improve the CAPM-SPECTRUM Integration to include automated lifecycle synchronization (add, retire, maintenance mode) - we have been working on this through a few releases now and in 2.6 we add Maintenance mode, with Spectrum 10.1 the team is adding the spectrum pieces to deliver this and then with CAPM 3.0 we tie the two systems together

Catalin Farcasanu to Everyone: Hi Seenu, Shiraz, glad to see you around here. Same for Daniel

Daniel Holmes to Everyone: @Catalin you just made my day !!!

Shiraz Ashraf to Everyone: Hey there Catalin

Seenu Mathew to Everyone: @Catalin. Always a pleasure to serve you. Thank you

John to Everyone: Thanks - eagerly awaiting 

Mike Collins to Everyone: @Sean on that is fine for timezone and not business hours but we have clients that require 8-18 hours dashboards and other dashboards with no business hours (one for client reports / dashboards and one for support purposes). So what what the avg cpu for my 8-18 BH and the same for my 24h BH and so on on many metrics , mostly on QOS/IPSLA ones were probes are in CA and other in Europe.

Alexander Bradley to Everyone: I had a situation in my lab environment where all three nodes in my Vertica database got wedged bad (it was a disk space issue)...to the point where I couldn't start it at all.  I was wondering if there is a way to dump the data and restart from an empty Vertica database while keeping the schema intact. I couldn't figure it out and ended up re-installing DR, DA, and the DCs instead.

Catalin Farcasanu to Everyone: @John: I managed to solve my problem, thanks to you and the last Office Hours meeting I'm thinking of other that might be in my situation.

Mike Collins to Everyone: @Sean does not really matter which timezone they are in if I say 8-18 then give me a dashboard with stats of the metrics I am asking for from 8-18

John Murdough to Everyone: @Alexander - Once the DB is wedged it's very difficult to clean up

Mike Collins to Everyone: Are there plans to actually move the datacollection failover (cannot communicate with the DA) stored in memory to disk anytime soon ?

John Murdough to Everyone: @Alexander - You can drop and recreate with admintools, but you would need the DA installer to rebuild in the DB, which is not ideal

Sean Moriarty to Everyone: @John - The time zone is needed on the device via site group otherwise not able to cross reference a given time value with start and end of a given business hour for that device.

Seenu Mathew to Everyone: @Mike Collins DC: DC can cache 45 min to unlimited of data during DA outage in 2.6 onwards VS 2.5 can hold only 15 to 30 min of data

james sabassi to Everyone: With the new CA app syntethic monitor, is it possible to schedule out maintenance periods for specific monitors (sites)?

Mike Collins to Everyone: @Seenu that is great

Mike Collins to Everyone: can you tell me how the DR's work ? is the data spread across the DR's or do they miror each other ? if one goes down do I have missing data on the PM ?

Catalin Farcasanu to Everyone: Are there any ways of removing a faulty certification from the system? I can always remove the configuration files from the DA, but the information store in the DR makes it difficult to avoid restoring of the DB. I've been there several times during the testing of certain device certifications. You can very easily add a new column, but you cannot change it's name for example or remove it.

Alexander Bradley to Everyone: BTW the DR installer on 2.6 works better than 2.5...thanks for the improvement!

Seenu Mathew to Everyone: @Mike Collins Thank you.

Sean Moriarty to Everyone: @John - So what what the avg cpu for my 8-18 BH and the same for my 24h BH and so on on many metrics , mostly on QOS/IPSLA ones were probes are in CA and other in Europe. Need to have further details and examples as I would receommend a formal ticket be submitted for further triage if the devices are properly configured. One caveat is parent device for the CPU needs to be within site group as a CPU itself in a device component and is not synchronized that must be doen via parent device.

Catalin Farcasanu to Everyone: Are there any plans to certify NQA from Huawei?

John Murdough to Everyone: @Mike - In a standard 3 node cluster there is 2 copies of the data, this is referred to as k-safety. So with a k-safety of 1 you can have a single node go down, and still be running.  Also with data segmentation you can have multiple nodes processing each query which improves performance

Mike Collins to Everyone: @John good I understand now thanks

Seenu Mathew to Everyone: @Catalin Feel free to raise a certification request through support for NQA from Huawei and we can get it certified

Daniel Holmes to Everyone: @James - can you confirm if you are speaking about the new ability to configure IPSLA tests with CAPM or one of our other data sources (i.e. ADA)

Mike Collins to Everyone: @Sean ok we can give you some examples but it is pretty basic to get all cpu from start to finish time according to business hours. but my question is what is the purpose of the timezone / business hour in the dashboard creation ? if I cannot pick my timezone and BH and get a table graph.

Melissa Potvin to Everyone: We are at the 1/2 hour mark - the team is looking for more questions, so give them some work to do - and ask away ...     we have support, engineering, services, product management all here!

james sabassi to Everyone: @Daniel - no, not IPSLA tests. This is using app synthetic monitor (formerly cloud monitor) to schedule maintenances so as not to skew availability reporting

Daniel Holmes to Everyone: @James I think you are speaking about a product within the APM Product Family but to be sure let me ask are you viewing the data within CA Performance Center? If not Melissa will help get you tied into the right community to help.

Sean Moriarty to Everyone: @John -Business hour filters are applied to a site group that has time zone. So is I have business hour 8am-4pm that is assigned to EST. However, I could use that  8am-4pm business filter on a colletion of devices that have been assoicated to PST within the view settings dialog.

james sabassi to Everyone: @Daniel This is where I view the data: https://asm.ca.com

Melissa Potvin to Everyone: @james: here is a link to the APM community: https://communities.ca.com/community/ca-apm

Sean Moriarty to Everyone: @John - Just that these devices that are in PST need to somehow be associated to that time zone via a site group assignment

Daniel Holmes to Everyone: @James yup that is the product within the APM Product Family - Melissa will help you

james sabassi to Everyone: @Daniel - thank you

Mike Collins to Everyone: I find that the features / roles that can be applied like Administer dashboards and so on are confusing. Because for example if I enable Proxy user feature but I don<t enable Administer users then the feature proxy users does nothing. Maybe regroup them in like if I give the administer users then I can give proxy users or not (but why was this done like it was from proxy users and not with add , delete or edit users) because we are looking to give our support team the ability to proxy to users but not add, delete or edit them.

Melissa Potvin to Everyone: @james; support teams highly responsive in APM community so you will be in great hands but reach out to me if you need anything

james sabassi to Everyone: @Melissa - Thank you, I will

Alexander Bradley to Everyone: I’m having trouble distinguishing between event rules definitions made in Threshold Profiles vs Monitoring Profiles.  In Data Aggregator it looks like there are two places to construct event rules, one in the Admin->Monitored Items Management->Manage Threshold Profiles, and the other  in Admin->Monitored Items Management->Monitoring Profiles.  Could you help clarify my understanding of these two event rules definitions?

Mike Collins to Everyone: @Sean why does this absolutely need to be attached to a time zone. Ex I have an IPSLA probe going from Montreal to Moscow , I put this probe in a site with Montreal timezone with 8-18h business hours, but what if in moscow the business hour is 3am to 9am and they want a 3 to 9 dashboard , but my device is already in a site group tagged with montreal timezone and 8-18h BH

Melanie Giuliani to Everyone: Hi all - this is the 15 minute warning! Please get your last minute questions/follow-ups in now.

John Murdough to Everyone: @Alexander - Threshold profiles provide you more flexiblity to target specific groups of component items or devices.  The rules that can be applied in Monitoring Profiles are applied to every component of the device that the MP is applied to.  Threshold Profiles were added more recenlty, but we chose to maintain the Monitoring Profile event rules in the product.  You're likely to find most use cases are better achieved with Threshold Profiles

Ernesto to Everyone: good morning all

have one manual of installation for dummies because have a lot problems with vertica db

Daniel Holmes to Everyone: @Ernesto - what version are you trying to install?

Mike Collins to Everyone: is there a possibility or plans in the near future to be able to use ODBC connections from outside sources to the database or not ? I don't think so but you never know

Ernesto to Everyone: @Daniel - 2.3.4

John Murdough to Everyone: @Alexander - Threshold profiles also can be created by users who don't have Data Aggregator admin privs, allowing you to give access to users to define thresholds without giving them polling and discovery control

Alexander Bradley to Everyone: @John Murdough thanks, John!

Daniel Holmes to Everyone: @Ernesto - we dramatically improved our docuementation focused on installation with CAPM 2.5 - you can now access these oneline at http://wiki.ca.com/capm - we've also been improving the installers to help make it easier

John Murdough to Everyone: @Mike - Although we are not planning to support direct DB access, we have released the OpenAPI for these types of use cases.  Have you looked at OpenAPI for your needs?

John to Everyone: Is there a way to better define the alerts from PC? The alerts from PC are sent to Spectrum (assigned to the model) most all of these alerts are "Performance Alerts" .. and suggest Launch the Performance View to see details... Regardless of "profile"  metric being reported...

Daniel Holmes to Everyone: @Mike our focus here has been to develop our OpenAPI to both make it easier to perform targeted data extractions but also to protect the integredy and performance of the system as you script and build apps against CAPM - it was release GA in 2.5 and some nice improvements in 2.6 based on early feedback from users

Ernesto to Everyone: OK thank you very much. I will check it

Mike Collins to Everyone: @Daniel we are working with the Odata and find that there is not a lot of documentation on creating searches (custom parameter), for example I have multitenants and would like to start my search or at least include the tenant name in my search and I am unable to find is there a detailed doc

Alexander Bradley to Everyone: Is there a way to delete a Data Collector from the Data Aggregator, then re-add it?

John Murdough to Everyone: @Alexander - Not specifically; however you can resinstall by setting the DCM_ID evn parameter and have the newly installed DC take over the workload.  For example if you wanted to move the DC to a new host

Catalin Farcasanu to Everyone: @Alexander: You need to install the DC having specified its DCID. That's the only way to install a new DC as an existing one. It's specified within the documentation. There's an environment variable that you need to set.

Daniel Holmes to Everyone: @John regarding your request to improve spectrum integration with more granular alerts defined - this is also under consideration for a future release in addition to the lifecycle synchronization I mentioned earlier - if you have an opportunity please take a few mins and search and vote, comment, etc. any stories related to spectrum integration on IM community (CA PM Subgroup) as this is an area we want to improve upon now that we are closing the gaps on core CAPM features

Sean Moriarty to Everyone: @mike – Proxy button operations provide proxy but not add/edit/delete. As far as I see in the code function in Menu Service does not consider if the user has the proxy users right when it looks to see if the User menu should appear, only the Administer users right. Should look up proxy and edit be considered within enable and disable. Right now we only consider administrator. This would be a short coming that should be addressed as a defect.

Melissa Potvin to Everyone: Last call for final questions. We have 5 minutes left - so please submit any final questions or thoughts.

Mike Collins to Everyone: @Sean how do I address this defect ?

John to Everyone: @Daniel - ty..

Daniel Holmes to Everyone: @Mike as a follow up to @Sean's explanation i will review with the product mgr and UX designer how we organize and present these admin features - as you shared we are adding more and more which is good but we want usability to enable you to take advantage and if there is an opportunity to improve we will

Mike Collins to Everyone: @Daniel thanks

John Murdough to Everyone: @Catalin - Relative to deletion of a bad cert.  We have plans to add a more specific capability in the future.  Today you can exclude the cert by removing it from the VC Priority list

Mike Collins to Everyone: That was great guys, got a lot of info from you guys.

Alexander Bradley to Everyone: yes, thanks all!

Daniel Holmes to Everyone: @Mike do you have a set of users you want to be able to proxy as other users but not manage (add, delete, edit) users?

Mike Collins to Everyone: @Daniel, yes

Daniel Holmes to Everyone: ok thanks!

Mike Collins to Everyone: @Daniel, my support team needs to be able to proxy to a user in a different domain though to be able to see what the client is seeing.

Daniel Holmes to Everyone: Thanks everyone for taking the time to join and ask great questions - 2.6 should be available any day and the team met yesterday to put the finishing touches on our 2.7 plan that we hope to release the end of this year and show at CAW!!!

Melanie Giuliani to Everyone: Alright, that's it for today. Thanks everyone for your great questions! We will be posting the transcript to the IM community later today.

Attachments

    Outcomes