We have a test enviroment to run VMmark3 benchmark. These are the versions of the components:
VMmark3 : vmmark3.1.1-template-031420
ESXi: VMware ESXi, 7.0.2, 17630552
We have successfully run a test with only first tile. After, we have deployed a second tile, again without errors. However, after the execution of a new test including the two tiles we get a "non compliant" result and we would like to know the reason.
We attach log file of running job
Thanks in advance
Hi @CSCM , unfortunately this STAX log is not enough to troubleshoot why the run failed.
Could you please zip and upload the contents of the primeclient results directory (/root/VMmark3/results/Results_20210913185815_Test2/) and post a link to it? Thanks!
Hi again @dmorse ,
In a deep inspection of log files we found this entry on WeathervaneCombined1.log file: (at workloadfiles folder)
Mon Sep 13 19:19:19 2021: Weathervane Version VMmark3.1
Mon Sep 13 19:19:19 2021: Command-line parameters:
Mon Sep 13 19:19:19 2021: showPeriodicOutput: 1
Mon Sep 13 19:19:19 2021: rampUp: 120
Mon Sep 13 19:19:19 2021: logLevel: 3
Mon Sep 13 19:19:19 2021: maxDuration: 10800
Mon Sep 13 19:19:19 2021: users: 9500,1500
Mon Sep 13 19:19:19 2021: steadyState: 10680
Mon Sep 13 19:19:19 2021: runProcedure: runOnly
Mon Sep 13 19:19:19 2021: Run Configuration has 2 workloads.
Mon Sep 13 19:19:59 2021: IP Address cannot be found for Client
Nevertheless, the WeathervaneCombined0.log is free of errors.
This is why we don't have the WeathervaneElastic1.wrf and WeathervaneAuction1.wrf files when we try to run
tilescore.pl -t 2 command.
Could you identify the VM that doesn´t have good hosts file?
Here are commands to zip up all the hosts files, which won't be very large at all. Please attach that to this thread.
zip hostfiles.zip *hosts.txt
For some reason, all of your tile1 VMs' hosts files are wrong: they are pointing to tile0 IP addresses. For example, AuctionLB1_hosts.txt shows me that for the AuctionLB1 VM, its /etc/hosts files are pointing to these tile0 VMs:
192.168.1.7 ElasticWebA0 ElasticWebA
192.168.1.1 Client0 Client
192.168.1.11 ElasticDB0 ElasticDB
192.168.1.2 DS3WebA0 DS3WebA
192.168.1.16 AuctionAppA0 AuctionAppA
192.168.1.14 AuctionWebA0 AuctionWebA
192.168.1.8 ElasticWebB0 ElasticWebB
192.168.1.9 ElasticAppA0 ElasticAppA
192.168.1.18 AuctionNoSQL0 AuctionNosql
192.168.1.13 AuctionMSQ0 AuctionMSQ
192.168.1.6 ElasticLB0 ElasticLB
192.168.1.3 DS3WebB0 DS3WebB
192.168.1.17 AuctionAppB0 AuctionAppB
192.168.1.5 DS3DB0 DS3DB
192.168.1.19 AuctionDB0 AuctionDB
192.168.1.15 AuctionWebB0 AuctionWebB
192.168.1.10 ElasticAppB0 ElasticAppB
192.168.1.4 DS3WebC0 DS3WebC
192.168.1.20 Standby0 Standby
While there are entries for Tile1 VMs at the bottom of the file, they don't have the correct aliases. I would have expected to see only these lines:
127.0.0.1 Client1.vmmark3.eng.vmware.com localhost localhost4 localhost4.localdomain4 Client1
192.168.1.28 ElasticWebB1 ElasticWebB
192.168.1.27 ElasticWebA1 ElasticWebA
192.168.1.26 ElasticLB1 ElasticLB
192.168.1.31 ElasticDB1 ElasticDB
192.168.1.38 AuctionNoSQL1 AuctionNosql
192.168.1.33 AuctionMSQ1 AuctionMSQ
192.168.1.21 Client1 Client
192.168.1.40 Standby1 Standby
192.168.1.24 DS3WebC1 DS3WebC
192.168.1.23 DS3WebB1 DS3WebB
192.168.1.22 DS3WebA1 DS3WebA
192.168.1.32 AuctionLB1 AuctionLB
192.168.1.39 AuctionDB1 AuctionDB
192.168.1.30 ElasticAppB1 ElasticAppB
192.168.1.29 ElasticAppA1 ElasticAppA
192.168.1.37 AuctionAppB1 AuctionAppB
192.168.1.36 AuctionAppA1 AuctionAppA
192.168.1.25 DS3DB1 DS3DB
192.168.1.35 AuctionWebB1 AuctionWebB
192.168.1.34 AuctionWebA1 AuctionWebA
Could you delete all the tile1 VMs (all VMs ending in 1, including Client1), and re-run the provisioning to re-create them? Hopefully then they will have valid /etc/hosts files that point to tile1 instead of tile0.
I´ve re-run the provisioning of entire tile 2.
All VMs are being deployed succesfully except ElasticAppA1 and ElasticAppB1.
I did a partial recreation of those two machines and I can see at event´s machine on vCenter than after being cloned from their respective machines of tile0, the process of customization fails.
I don´t see root cause on log located at /var/log/vmware-imc/toolsDeployPkg.log in the guest OS.
Thanks in advance
Something is definitely wrong -- there should not be any errors when creating tile1 VMs, and you should not have to do a partial creation of ElasticApp VMs. Please do the following: