Thank you for joining askCA for CA Performance Management today, January 18. Great participation and questions!
What follows are the transcripts from today's session:
------------------------------
from Melissa Potvin (CA) to everyone: hi everyone, let's get started and welcome to askCA for CA Performance Management!
from Melissa Potvin (CA) to everyone: i have support engineers here with me and product management. go ahead and ask your questions!
from Francisco Ramilo to everyone: Good afternoon all
from Francisco Ramilo to everyone: CA PM v3.2
from Francisco Ramilo to everyone: upgrade to CA PM v3.5 any major issues found in your working experience
from Francisco Ramilo to everyone: or is it a "smooth" upgrade?
from Andrew Cuthbertson to everyone: upgrade worked fine
from Jason Normandin to everyone: @Francisco - The CAPM 3.5 upgrade should be very straight forward. There is no Vertica upgrade required and we had over 20 customers participate in our 3.5 beta. We did not find any issues related to upgrade that should cause concern.
from rjohn to everyone: Hi! Thx for taking questions. In CAPC, is it possible for a user to get the required ADA-datasourced report that can be found in ADA->Engineering tab using a filter in CAPC which is like the Settings filter which is in ADA? If not, is that something you would consider implementing in CAPC?
from Francisco Ramilo to everyone: @Jason thank you!
from rjohn to everyone: Basically, the ability to filter by Server/Application/Network to get required ADA datasourced-report in CAPC
from Jason Normandin to everyone: Hi @rjohn - Using CAPC grouping structures should provide some level of seperation but specific per view filtering is only possible via groups, items, and time. You would need to use the dashboard context selector
from rjohn to everyone: is the context selector a new feature in 3.5? For better network device definition?
from Jason Normandin to everyone: @rjohn - No. This is the group filter within CAPC that controls what is shown in each Dashboard.
from rjohn to everyone: ok, thx
from Jason Normandin to everyone: We do have very rich grouping capabilities however that will allow you to control how those groups are created using ad-hoc or rule based rule definitions
from rjohn to everyone: well, its for troubleshooting
from rjohn to everyone: so pre-creation of groups is not a good option
from Paul Burland to everyone: Hi, I'm hoping to get involved with an intergration project, but understand that the CAPM API has room for improvments. is there a time line for making the API better?
from Andrew Cuthbertson to everyone: metric families: where (ie the files) can I find the sysoid that is used to determines whether a metric family is supported by a device. (aside comment - would be nice to have util to test which families are supported so you don't have to implement a monitoring policy to find out.)
from bjorn to everyone: Hi! A question about timestamp on alerts.
What is the source for timestamp on CA PM alerts ?
Looks like alerts are timestamped with the poll interval. Is this correct ?
If so. this looks strange when forwarded to spectrum.
The doc says that with 60 sec event polling interval from spectrum gives a 1-2 min delay before the CA PM alert appears in spectrum.
But have lots of incidents where eg CA PM time stamp are 5:50 and Spectrum 5.54, or PM 2.45 and Spectrum 2.49. It will say 4 min delay.
from Jason Normandin to everyone: @Paul - Can you be more specific as to the limitations? We have many customers using our API for many types of integration, data extraction, data visualization, and administration functions
from Jeffrey Pinard to everyone: @Andrew, there is no tool. we have the http://DA:8581/typecatalog/certifications/snmp
that list the makeup of the Vendor Certs. This includes the OIDs each VC can read. Each OID has a IsKey. When true, we use them for VC supported or not.
from Jeffrey Pinard to everyone: @Andrew, these files are stored in /opt/IMDataAggregator/apache-karaf-2.4.3/certifications/CA/vendor_certs in 3.2 and older and /opt/IMDataAggregator/data/certifications/CA/vendor_certs in 3.5+
from Nick Brown to everyone: Hi @Jason, In 3.5 it appears more CABI features are added, Does CABI reporting/data extracts come direct from vertica or is it using OAPI?
from Andrew Cuthbertson to everyone: to diagnose snmp request/reply issues - where are the best places to look. eg I was trying to find why no interface byte values in db so wanted to see if any were being requested/returned from devices.
from Jason Normandin to everyone: @Andrew - We also have an OpenAPI App that can provide a lot of this info for already discovered devices. Not a direct solution to your problem but helpful to understand what a monitored device is supporting. https://github.com/CA-PM/DeviceSupportedMetrics
from Jason Normandin to everyone: @Nick - OpenAPI
from Michael Poller to everyone: @bjorn, sorry for the delay as we discussed your timing question re: Spectrum<->CAPC alarm timing. Assuming the servers are time synced using the same time server there shouldn't be that large of a gap.
from Michael Poller to everyone: @bjorn I'd recommend a new support case to investigate the lag and time delay being observed.
from Jeffrey Pinard to everyone: @bjorn, what is your event polling rate in the Spectrum Performance Center settings? 1 min (OOTB) or greator?
from Paul Burland to everyone: @Jason, I haven’t got involved as yet so don't have specifics. Really just a general question on API development. It's around provisioning in multi Tennent environment
from Michael Poller to everyone: @Andrew Cuthbertson, the first place we'd look to diagnose problems like that is our 'DCDebug' page.
from Jason Normandin to everyone: @Paul - We have a lot of admin API's that can be driven to manage and deploy tenants.
from Aravind Sugur to everyone: @Andrew - you can use dcdebug -DA:8581/dcdebug/searchdebug.html
from bjorn to everyone: Polling rate are 1 min, and it looks like all CA PM events comes in 5 min intervall 0,05,10,15 and so on
from Michael Poller to everyone: @Andrew Cuthbertson; it is found at the URL http://DA_HOST:8581/dcdebug/searchdebug.html.
from Michael Poller to everyone: @Andrew Cuthbertson; can use it to review polling configurations, polling errors, discover logging messages and can also enable poll logging and snmp logging to examine snmp request/response pairs during poll cycles for a given device
from Nick Brown to everyone: @Jason is the CABI data feed via OpenAPI any more or less efficient then using the OpenAPI to extract data via a script?
from rjohn to everyone: For OpenAPI access, is it possible to convert http://DA:8581/odata/api to https to that SSO login credenitials that are entered by the user is not sent clear-text?
from Jason Normandin to everyone: @Nick - The CABI data feed will not provide the raw data - it's designed as a presentation tool, not an extraction tool. Going direct to OpenAPI will get the backing data without presentation
from Andrew Cuthbertson to everyone: custom attributes: can be populated via REST. Any plans to allow entry via GUI? eg details dashboard view in device/interface context pages
from bjorn to everyone: About alert text. Is it possible to change alarm text and parameters for an alarm sent to Spectrum. We want ifalias and
the actual value at the time when a threshold is crossed, in the alarm text or as parameters in the alert sent to Spectrum.
from Jason Normandin to everyone: @Andrew - There's an App for that - https://github.com/CA-PM/Device-DA-Admin
We are also looking to continue to evolve item management via CAPC UI in coming releases - For example, we added the ability to define alias, speed, and IP info in 3.2 via CAPC
from Jason Normandin to everyone: @Andrew - This app can help as well: https://github.com/CA-PM/DA-Utility-Apps
from Jeffrey Pinard to everyone: @bjorn, we only send the event and it's properties to Spectrum when sending over the event. The Event/Pcause files for Threshold Violation on the Spectrum side can be modified to add more info, but not sure how to get access to ifAlias.
from Nick Brown to everyone: @Jason - CABI for presentation is it extracting the realtime as the person runs the report? Does it allow for Non-Business Hour, Business Hour, and All Hour based on specific TimeZone of the device/component location?
from Joe Rossi to everyone: @rjohn: If you are making direct OpenAPI calls, then you may want to use the OData Proxy on the PC server. See: https://docops.ca.com/ca-performance-management/3-5/en/using/openapi/openapi-apps
this mentions it. Assuming PC is enabled for SSL, any OpenAPI call can be made via https://pc:8182/pc/odata/api This will proxy to DA (which our best practices say should be behind a fire wall), but since PC is setup for SSL, the creneditals will be encrypted from the browser to the server.
from rjohn to everyone: ok, thx...
from Joe Rossi to everyone: @rjohn: For our release we are just starting, we are looking at enabling the ability to configure the Data Aggragator to use SSL. I'd watch the sprint demos to check the progress of that.
from Jason Normandin to everyone: @Nick - We support these at the report level and not at the individual device or component level - Basically you can chose the BH and TZ for the report and all items in that report will share the same BH and TZ.
from rjohn to everyone: ok, thats good
from bjorn to everyone: ifalias is a attribute on the device in CA PM, so it would be nice to send this with the alert, an not only eg gi0/1
from LeRoy to everyone: When upgrading from 3.2 to 3.5 can I use the same response file from 3.2 with the 3.5 install?
from Jason Normandin to everyone: @bjorn - We have a feature in an upcoming release that will allow you to run an action script from an event. In this script we will provide a lot more attributes and info but you can also easily kick off an OpenAPI query to get additional config info (or data) and then send that alarm over to Spectrum using our TrapExploder utility -
from Michael Poller to everyone: @LeRoy, as long as the values/details are the same for the server in the existing response file, as long as the expected values per the docs are present, I don't see why it can't be used.
from Michael Poller to everyone: @LeRoy; here is the URL for the r3.5 documentation re: creating a response file. https://docops.ca.com/ca-performance-management/3-5/en/installing/install-the-data-aggregator
from LeRoy to everyone: That talks about the DA. I'm asking can the same file be used for both 3.2 and 3.5?
from rjohn to everyone: With the new features for customization of UI (and this may already be possible), can I upload company logos/image file to the dashboard?
from bjorn to everyone: And when we get a threshold violation, you get that its eg ">0" but not the value. It would be nice to get the value at the time when the threshold is broken
from Jeffrey Pinard to everyone: @LeRoy, in 3.5 we added HA DA ability, so there are some more questions that need to be in the response file.
from Melissa Potvin (CA) to everyone: please visit Product Roadmaps & Feedback Sessions - CA Technologies to get registered for a product roadmap and feedback session. we just did one (yesterday) for PM but more are planned for Feb and march. registration is required ,,,,
from Aravind Sugur to everyone: @LeRoy I dont see a reason why it would not work for DA . Any issues please feel free to log a support ticket
from Nick Brown to everyone: @Jason - OpenAPI do you have any customers running OpenAPI queries in parallel to help extract bulk data,ie. 500k-1m records? if so, how is it working for them?
from Ernesto to everyone: can be installed on linux 6.8 CA PM 3.5
from Jason Normandin to everyone: @Nick - Yes! We have customers doing very large extractions. OpenAPI supports many modes of data retrieval including paging for very large results sets.
from Michael Poller to everyone: @LeRoy; The PC is reusable but best to regen the DC and DA ones due to changes (FT proxies and other new options)
from LeRoy to everyone: Thanks
from Jason Normandin to everyone: @rjohn - Yes! With the theme support added in 3.5 you can change colors, fonts, images (including logs) etc. The copyright and the chart colors cannot be modified but we are working on a project to support custom chart colors as well.
from Aravind Sugur to everyone: @Ernesto Yes you can install CAPM 3.5 on 6.8 - Please refer to https://docops.ca.com/ca-performance-management/3-5/en/installing/review-installation-requirements-and-considerations
from Andrew Cuthbertson to everyone: are you checking cabi on rh7.4 so evrything relating to v3.5 capm can be on the same os version
from Paul Kelly to everyone: @Melissa; My browser threw a fit and I lost the full thread of this session, would it be possible for you to email it to me please? There are bits I'd like to extract.
from Michael Poller to everyone: @Paul Kelly; Yes, Melissa will post the chat transcript to the public community pages after the session.
from Melissa Potvin (CA) to everyone: @paul, michael is correct. check back in the community, transcript will publish later today.
from Ernesto to everyone: another server is required for CA DA proxy
from rjohn to everyone: Is it possible to get the device Last Discovered and/or Last Modified date from the REST API?
from Michael Poller to everyone: @Ernesto; correct the new FT DA Proxy configuration does require new servers
from Jeffrey Pinard to everyone: @rjohn, we only have when it was first discovered via REST in CreateTime value.
from Jason Normandin to everyone: @rjohn - You can get CreateTime from the OpenAPI for any device or component but I'm not sure about last modified time
from Aravind Sugur to everyone: @Andrew CABI is supported on 7.2
from rjohn to everyone: I'm asking so that I can create a script to auto-delete old devices (say not discovered in the last 6months) - it would be good if DA could have this feature built-in
from Aravind Sugur to everyone: @Andrew Please refer to https://docops.ca.com/ca-business-intelligence/6-3-0/en/installing-ca-business-intelligence-jasperreports-server
from rjohn to everyone: so you see create time wont help with what I am trying to do
from Jeffrey Pinard to everyone: @rjohn, do you mean things that haven't been polled in last 6 months?
from Melissa Potvin (CA) to everyone: 5 minute warning. please submit final questions
from rjohn to everyone: yes, that will work if not polled means that it is not pollable
from Nick Brown to everyone: @Melissa do you have a url for the 3.5 features presented last week?
from rjohn to everyone: status?
from rjohn to everyone: but status does not help because it does not have a date
from Melissa Potvin (CA) to everyone: @nick, here you go.. https://communities.ca.com/videos/6443-webcast-replay-what-s-new-in-ca-performance-management-35
from Jason Normandin to everyone: @rjohn - You can key off of the Reachability metric family to view devices not reachable for a specific time interval. I think we have this documented in the OpenAPI - advanced examples doc area
from rjohn to everyone: ok, thx - will take a look
from Jason Normandin to everyone: If you can't find it, shoot me an email: jason.normandin@ca.com and I'll be happy to help.
from rjohn to everyone: thx!
from Melissa Potvin (CA) to everyone: OK everyone that is a wrap for today. GREAT questions!!! thank you for joining us today and we hope you found spending your time here useful.