from DK to Everyone:
g'morning Lenn and Al
from Alfred Polanco to Everyone:
@DK good morning
from Alfred Polanco to Everyone:
@DK yes. Deekshit. that name rings a bell. glad to have you in the chat.
from Lenn Thompson (CA) to Everyone:
@Everyone: Leg's go ahead and get started. What questions do you have this month?
from Jim Spath to Everyone:
Hello all.
from Lenn Thompson (CA) to Everyone:
@Jim Good morning, sir. Thanks for joining.
from Alfred Polanco to Everyone:
@Jim Spath: Welcome, Jim
from DK to Everyone:
I did have a question about certifying Autosys to work with SQL Server on Linux, now that MS supports on Linux
from DK to Everyone:
a client of mine has been asking for long and I wanted to make sure they were heard..
from Alfred Polanco to Everyone:
@DK: you may need to create an ideation for this
from Jim Spath to Everyone:
I've gone through installing 11.3.6 SP7 to a new Windows OS, including EEM, WAAE and finally WCC. Whew.
from DK to Everyone:
Thanks Al. I will
from Jim Spath to Everyone:
Primary general question is around statistics aggregation.
from Jim Spath to Everyone:
It looks like stats collection is now the default.
from Jim Spath to Everyone:
Was "AggregateStatistics=0", now "AggregateStatistics=1" in the config.ACE file.
from Alfred Polanco to Everyone:
@Jim Spath: Yes. That is true:
from Alfred Polanco to Everyone:
https://docops.ca.com/ca-workload-automation-ae/11-4-2/en/release-notes/ae-release-notes/ae-release-11-3-6-sp1/changes-to-existing-features-11-3-6-sp1
from Jim Spath to Everyone:
There seems to be only "delete all" option. We will need to monitor space growth when this gets to production.
from Jim Spath to Everyone:
I plan to turn the feature on before the next upgrades. Any gotchas to be aware of? I see the manual says run manual collection first before automating.
from Alfred Polanco to Everyone:
@Jim Spath: We support deleting partial statistics:
from Alfred Polanco to Everyone:
https://docops.ca.com/ca-workload-automation-ae/11-4-2/en/administrating/ae-configuration/retrieve-aggregated-job-alarm-and-scheduler-statistics/delete-aggregated-statistics
from Jim Spath to Everyone:
Oh, good.
from Alfred Polanco to Everyone:
@Jim Spath: I am wrong. It does appear to be an all or nothing deal
from Jim Spath to Everyone:
Probably have 20 million records in ujo_proc_event. Any ball park on how long stats create will take? Manual says "some time".
from Jim Spath to Everyone:
@Al I can open a ticket on partial removal. I'd guess no referential integrity to consider.
from Jim Spath to Everyone:
The other "new" feature I need to get up to speed on is the web service required for agent inventory. Our last install ran into conflicts and the web server would not stay running. Now looks like we'll need it.
from Alfred Polanco to Everyone:
@Jim Spath: We will also have to collect statistics about how long it takes to collect statistics.
from Jim Spath to Everyone:
@Al :-)
from Jim Spath to Everyone:
I will let you know how long catch-up takes in development once the DBAs are informed. I will propose putting the ujo_rep* tables into their own tablespace for space management.
from Alfred Polanco to Everyone:
@Jim: Please do. Thank you.
from Jim Spath to Everyone:
For WCC 11.4, I had to create different Oracle DB user ID as the preexisting tables conflicted with new versions. Seems to have worked.
from Alfred Polanco to Everyone:
@Jim: Interesting if you could provide more information to the WCC team via a support ticket, it can help us improve the upgrade experience in a future release.
from Jim Spath to Everyone:
One continuing pain point is around Windows drives with "8dot3" naming. WCC would not install to "E:\Program Files..." at all. I ended up using "E:\WCC".
from Jim Spath to Everyone:
@AL OK. I have an internal document I'm creating to guide the next upgrades. Due to policy, new Window OS is not an upgrade but a reinstall to need to start from scratch. I was able to reconnect to the earlier AutoSys DB with minimal fuss. So that's a plus.
from DK to Everyone:
We see a lot of errors about IOExceptions in our WCC logs in prod...we dont see them in prod
from DK to Everyone:
Caused by: java.io.IOException: An existing connection was forcibly closed by the remote host
INFO | jvm 1 | 2018/04/08 11:32:29 | 11401 | at sun.nio.ch.SocketDispatcher.write0(Native Method)
INFO | jvm 1 | 2018/04/08 11:32:29 | 11401 | at sun.nio.ch.SocketDispatcher.write(SocketDispatcher.java:51)
from DK to Everyone:
What would be the cause of these issues? We didnt see anything break in WCC, but we are concerned there could be something we are overlooking
from DK to Everyone:
sorry, we see them in our prod wcc logs, not in DEV
from DK to Everyone:
this is WCC 11.4 SP5
from DK to Everyone:
org.apache.catalina.connector.ClientAbortException: java.io.IOException: An existing connection was forcibly closed by the remote host
from DK to Everyone:
is it just Tomcat cleaning up its connection pool? or is it something more dangerous that needs to be fixed?
from Jim Spath to Everyone:
I was unable to find usefu 11.3.6 sp7 email template examples in the ops docs. References are scattered on multiple pages (e.g. https://docops.ca.com/ca-workload-automation-ae/11-4-2/en/administrating/ae-configuration/send-email-notifications )
from Jim Spath to Everyone:
I have not fully tested email but this statement is worrisome - "CA Workload Automation AE does not send the email notification even if one of the multiple email addresses you specified is invalid."
from DK to Everyone:
+1 @Jim..same here..its not in one place...I had a call with Sumitha from the docs team and they promised to organize it better
from Alfred Polanco to Everyone:
@DK: You would have to open a ticket for WCC but on the surface it sounds like HTTP client has a persistent connection that the server is closing perhaps as part of some internal maintenance
from Jim Spath to Everyone:
@DK I was able to piece it together but yeah, multiple examples with different scenarios could help.
from DK to Everyone:
I think the docops site is not something that can be easily navigated...I give up most of the times and just download the PDF and use the Ctrl+F to search for what I want
from DK to Everyone:
@Al - I will open a ticket
from DK to Everyone:
with WCC, and IE11 (on Windows 10) we have been seeing ECMCA script errors..
from DK to Everyone:
@Jim have you seen these errors? I posted a screenshot on the community and opened a ticket with CA too
from Jim Spath to Everyone:
Back to stats collection... Example code looks fcused on hourly or daily patterns. But the common bottleneck is many jobs scheduled to start at "the top of the hour". So I would like to pull information at the second or minute timeframe for those busy times. Aggregated stats are great for capacity planning, not so good for performance tuning. Sys admins say "your average CPU is low" but it's the peaks that hurt.
from Shannon Hebert to Everyone:
Hi DK, we're aware of this and working on a fix for the WCC in IE
from Jim Spath to Everyone:
@DK I try not to look at java error logs if I can avoid it. I never know which are useful and which are noise. I'll check on those if you post a general question on the community site?
from DK to Everyone:
Thanks Shannon! thats unfortunately not what the support person is telling me, but I think we are on the right track if there;s going to be a fix
from Shannon Hebert to Everyone:
DK, what's the case number and I'll review
from DK to Everyone:
00999174
from Jim Spath to Everyone:
I mostly like the new WCC flow view. I've got it working in the sandbox and am waiting for a change to preview with the scheduler team.
from Jim Spath to Everyone:
It's not quite JobScape. :-)
from Alfred Polanco to Everyone:
@Jim: blast form the past. love it. :)
from Jim Spath to Everyone:
Last comment... I have run into several issues around spaces and Windows commands after 11.3.6 SP7 agent installs. One is on DBMaint path being set as %autosys%... without quotes. Also Windows could not find the command "REN". That was a head scratcher!
from Alfred Polanco to Everyone:
@Jim: The DBMaint issue sounds like a known problem:
from Alfred Polanco to Everyone:
https://support.ca.com/us/download-center/problem-detail.html?docid=654169&productcd=ATSYS&problemnbr=4714
from Jim Spath to Everyone:
@Al probably because I hit it the last time...
from Alfred Polanco to Everyone:
@Jim: this was a fairly recent fix and may still be undergoing verification from qa
from Alfred Polanco to Everyone:
@Jim: OK. So there is movement on your issue
from Jim Spath to Everyone:
@Al We worked around by running DBMaint from a scheduled job.
from Alfred Polanco to Everyone:
@Jim: Understood.
from Jim Spath to Everyone:
The old ticket was - CA Support Request 21748541 01 - DBMAINT NOT RUNNING ON WINDOWS
from Jim Spath to Everyone:
Thanks all. Till next time!
from DK to Everyone:
Will AEWS support job log retrieval?
from DK to Everyone:
std_err, std_out
from DK to Everyone:
there are a lot of teams that use AEWS and embed the ret client calls into their logic
from DK to Everyone:
rest client calls *
from Alfred Polanco to Everyone:
@DK: if it is not a documented as part of the data schema, then you need to create an ideation
from DK to Everyone:
sure...its just hard to know what CA is thinking when we create an Ideation specially if CA doesnt respond...
from DK to Everyone:
:-)
from DK to Everyone:
@Al is there currently an autoping API or Webservices?
from Alfred Polanco to Everyone:
@DK: The AEWS schema is in:
from Alfred Polanco to Everyone:
https://docops.ca.com/ca-workload-automation-ae/11-4-2/en/reference/ae-web-services/data-schema
from Alfred Polanco to Everyone:
@DK: I don't see that there is an agent ping. Let me check our SDK docuemntation.
from DK to Everyone:
ok no autoping call...
from Lenn Thompson (CA) to Everyone:
@Everyone: We have a few minutes left. What else would you like to cover today?
from Alfred Polanco to Everyone:
i don't see an autoping feature in the SDK either
from DK to Everyone:
yeah..
from DK to Everyone:
unfortunately :-(
from DK to Everyone:
so I was testing the email template options...and the blob filename always shows up wrong
from DK to Everyone:
autorep -z ALL
Glob Name File Name
________________________________________________________________ ________________
emailTemplate2 /tmp/emailTemplate2
from DK to Everyone:
always as /tmp/emailTemplateblah instead of the actual path
from DK to Everyone:
I think this issue has existed since r11 though
from DK to Everyone:
since whenever blobs and globs came into existence
from DK to Everyone:
is this a known issue or is there a planned fix for it soon
from DK to Everyone:
?
from Lenn Thompson (CA) to Everyone:
@Everyone: Okay, folks. We'll have to wrap up for this session. DK, please feel free to ask those questions out in the community.
from DK to Everyone:
the issue with the blob too? I suspect thats an issue...I will ask in the Community too