pcampbel
Contributor
Contributor

Re-enable monitoring

I had disabled and removed most collection, as we only need to see specific information. Now, however, someone wishes to see an extra piece, filesystem % utilization. I have went into Administration - Monitoring Defaults - FileServer Mount and re-enabled collection and indicator.
Select Inventory under the platform in question, Services, but nothing exists there. I stopped the hq-agent, cleared out /install-dir/data directory, restarted the agent, also tried running setup, nothing. Went to the Monitoring section for the platform, selected the Metric's subtab, Show all Metrics......data I am looking for is not there.

Do I have to recreate by hand, all the FileServer Mount information to monitor % used or am I missing something to get that information back into a monitored state.

tia
0 Kudos
14 Replies
excowboy
Virtuoso
Virtuoso

Hi pcampbel,

Go to your service Monitor page. Select "Metric Data". Check the boxes in front of the values you would like to reenable for monitoring, click on the button "Show all Metrics" fill out "Collection Interval for Selected:" and klick the button with the arrow right next to it.
I think this is also mentioned in the review guide.

Message was edited by: excowboy
0 Kudos
pcampbel
Contributor
Contributor

That would be the problem, they do not show up. I have enable the metric data for them globally, but they do not show up under the individual platforms....anywhere that I can find.
0 Kudos
excowboy
Virtuoso
Virtuoso

Go to the Monitor-Page (where you see the indicators) of your platform/server/service and then "Metric Data" and not to Administration -> Monitoring Defaults.
0 Kudos
excowboy
Virtuoso
Virtuoso

I forgot to mention that you have to click on "Show All Metrics:" when you're on the metric data page.
0 Kudos
pcampbel
Contributor
Contributor

Correct me if I am wrong, I think you are describing going to "Browse Resources" click on the platform that you wish to see the data on, which puts you on the Monitor page, then select the "METRIC DATA" subtab, then the right arrow next to "Show All Metrics".
This does list several things but it does not show the Disk usage statistics, it shows data on cpu, memory, load average, number of cpus, number of running processes, and swap.
I don't see any indication of multiple pages either.
0 Kudos
pcampbel
Contributor
Contributor

I've tried everything suggested so far but have not been able to re-establish monitoring of the disk statistics. Anyone have any ideas as to how to re-establish this.
TIA
0 Kudos
excowboy
Virtuoso
Virtuoso

Okay, now I see the problem: You have deleted your file FileServer Mount Services and not only several metrics ? I think you have to add them manually.Maybe running Autoinventory from commandline again is an option too:
http://support.hyperic.com/confluence/display/DOCSHQ30/Autoinventory+CLI+Command

Message was edited by: excowboy
0 Kudos
pcampbel
Contributor
Contributor

Ouch, I was hoping for a better answer, kind of expected this answer but.....
Guess I won't be monitoring anything but Availability for a while, at least until I can figure out a better way.
Thanks
0 Kudos
admin
Immortal
Immortal

Did you really remove your file server mounts from HQ? I'm assuming they are still there. Go to Browse Resources->Services, then select FileServer Mount from the dropdown. How many do you have in your inventory?

If you tried to re-enable these metrics from the Administration section it's possible the operation is failing due to the large number of these services you may have in your inventory. I would try re-enabling the metrics again, then check the server log for any errors.

-Ryan
0 Kudos
pcampbel
Contributor
Contributor

87 show up under FileServer Mount, from just a few servers.
The only error I see is ERROR: duplicate key violates unique constraint "eam_measurement_data_pkey"
0 Kudos
pcampbel
Contributor
Contributor

I also see these in the server.log

INFO [org.hyperic.hq.appdef.server.session.AIQSynchronizer] Platform 'platformname' has ignored servers, leaving in queue.

ERROR [org.hyperic.hq.bizapp.server.session.AppdefBossEJBImpl] Invalid entity still in resource table: 3:19483
0 Kudos
pcampbel
Contributor
Contributor

Would it be possible to export the current data from the DB, reinstall the server piece and then re-import the data into the new clean database?
Or would another solution be better?
0 Kudos
pcampbel
Contributor
Contributor

May have a way to fix this but will need some assistance.
I have an April copy of the DB. I would think I could make a copy of the current install, backup the current install DB, move the old DB copy into place and restore the current data back into it. How to do this cleanly? I tried on my own, in a test instance, did a bin/db-psql.sh<backup, to recover, received lots of errors about violating constraints, etc. The db restarted okay but it doesn't look like the data repopulated the db.

TIA
0 Kudos
pcampbel
Contributor
Contributor

By doing some digging I have come up with at least a partial answer for the missing FileServer Mount data.
Hopefully someone can point out the error of my ways, please, or confirm that this should work.
What I did was compare the table sizes against an old backup that still had the now missing data.
The following tables looked the most promising as the most closely changed in size and containing the data I wanted to recover. By changed in size I mean that I went by the fact that I had removed approximately 7000 FileSystem Mounts.
I exported these tables from the old backup

eam_config_response
eam_service
eam_server ( not sure if I needed this one )
eam_resource
eam_res_grp_res_map

and then imported them into a copy of the current production db, in the order listed, and I appear to now have all the FileSystem Mount stuff back.
I have pointed one server to this recovered instance, and it seems to be gathering the FileSystem Mount data nicely.

Can anybody see anything that I have missed.

TIA
0 Kudos