How to install Hitachi Content Monitor on a single VMware instance.
Hi Thomas, graet work but I have one issue. I have followed your instructions and got into HCI deployment. After a reboot the HCI.service never starts again and the errror msg is: "HCI.service: Failed at step EXEC spawning /opt/hci/bin/run: Permission denied". Since I´m not really familiar with troubleshooting the linux environment I would like to ask if you have an idea?
Thank you Ralf,
Unfortunately I do not know the answer to your problem.
Could you please post it as a question in the forum, so someone smarter than me can help?
I will do that.
Would be best if you posted logs/screenshots.
1. Are you running HCI from systemctl or from the shell directly?
2. If directly what user are you logged in as?
3. If not 'root' are you using 'sudo'?
4. Where did you install HCI, i.e. what is the install directory?
5. What OS are you using? What version of docker? Are you running w/SELinux enabled?
I guess the docker version was not 1.12.6 but a bit higher. I have restarted the complete installation yesterday and today the deployment has finished and even after reboot of the OS the HCI.service came up again. I did all steps like in the video (user root, install directories etc. ). Next time I run into an issue I will put logs and/or screenshots in the post.
Thx for ur help.Ralf
Jon, any idea on my issue with Anywhere - silent registration (post in the HCP Anywhere developper group)
Thats a "diffcult one" and I hope for my customer that we get it running ...
So it is working, yes?
I created the monitor, connected successfully to my HCP but no data is showing.
All VMs (HCP, HCI, W2012 etc.) are running on my laptop. So ressources are limited. I gave my HCI instance 1 core and 16GB of memory. Can this be a reason for not filling up with data?
HCI needs a minimum of 4 cores and 16GB to run. Yes, this is HCM, but starving the CPU resources on a box that needs them is never a good idea. How much RAM/CPU does your laptop have? I would not be surprised if HCI and HCP are simply fighting to co-exist.
Are there any alerts, services not running?
What version of HCP? 7.3 or 8.0? If you are running 7.3 this dashboard will be empty b/c it is populated by node_status and this API wasn't available until 8.0. Try the storage and objects dashboard.
my laptop has 2cores and 32GB RAM
There are no alerts and it seems that all services are running. HCP VM is installed in version 8 with 1 core and 8g RAM
no dashboard has data
syslog, snmp, MAPI, node status, are configured as in the video. Is there an option on the HCI/Fedora CLI to check if data can be collected from the HCP?
syslog is PUSH from HCP to HCM. To verify that you are receiving data on syslog you can set up a syslog server on the HCM node (it would be listening on the normal 514 port). The other 3 are PULL signals and HCM is making API calls to HCP. you can execute the same calls with curl and snmpwalk. Again, do this from the HCM node.
There are log files on HCM: <installdir>/log/com.hds.ensemble.plugis.search.monitorApp.
Go into the Admin App and check monitor workflow to see if it is stopped/halted for some reason. Then also check the index collections to see if they have any documents.
I will check ...
Check SNMP, make sure the community matches on HCP and HCM. I use public.
In syslog make sure you put port 9601 at the end of the IP address of the syslog server (hcm ip)
Hi Thomas, hi Jon,
I gave a second core to my VM and now (though I have several VMs running on my laptop and struggeling for cpu/memory ressources) the dashboards have been filling up with data:
Thanks for your help and patience. I will now "play" va little bit around it.
Please look at my reponse to Thomas
HCI and HCM are not designed today to run on skinny machines, or fat ones that have tons of stuff running on them. We are working on slimming down HCI. That is part of our Foundry Decoupling effort.
It worked for me from the first try. I setup HCM VM with 4 CPUs and 16GB of RAM on ESXi server (same where 3-nodes HCP-VM are running). It took about 3 min to populate data for all Dashboards.
Retrieving data ...