Hi all,
We are currently doing a clean up of (ghost sites) gather errors on some of our relays. We have matched the site ids and site names through the Mirror dir, gathertate and bfe mapfile xml files from the relays and identified through the relay logs which endpoints are requesting the Ghost sites from the relay.
Sample errors on the relay logfile.txt so you get what I mean:
Unable to find ( or problems occurred ) with site id for URL: ``http://FQDN:52311/cgi-bin/bfgather.exe/opsitenumber`` - from (ENDPOINT IP)
At first, we just simply did a gather state reset, but we noticed the site errors are present again, thus the need to hunt down the endpoints that are still requesting the ghost sites. (Mostly Opsites)
We created a custom script to delete the contents of the BESData dir on the endpoints requesting the ghost sites and will do a Gather state reset on the parent relay.
For anyone who has done this kind of clean up on their environments, are we doing the correct procedure? Or is there a better way to handle this?
We understand that the gather errors are background noise and do not cause any real issues atm. We just want the site gather errors cleared. Thanks all we would appreciate any tips and tricks.
Run the following command from the Root Server's command line:
besadmin.exe /propagateAllOperatorSite
Let us know if this fixes the issue or not. The issue is not with relay though here
Hi … rather than the propagateAllOperatorSite is an admin tool command a little bit heavy and is better to be used when other attempts had no success, please better run from the server using a MasterOperator account the following:
a. iem login --server=<hostname>:<port> --user=<username> --password=<password>
b. iem get operators > operators.xml
c. iem post operators.xml admin/propagate/operators
About the IEM command line utility:
https://developer.bigfix.com/rest-api/iem_cli/
1 Like
We will try this and will let you know how it goes. Thank you.
How about if we have the same errors with a custom site?
From relay logfile.txt:
Date and time +0000 - 5520 - Unable to find ( or problems occurred ) with site id for URL: http://RootFQDN:52311/cgi-bin/bfgather.exe/CustomSiteName - from (ENDPOINT IP)
Date and time +0000 - 5520 - 123: class NotASignedMessage - from (ENDPOINT IP)
Could try the following procedure, works for existing custom sites:
-
Stop all BigFix services
-
Run the following SQL update once connected to the database for the server
use BFEnterprise
insert into dbo.PENDING_PROPAGATIONS (SiteType, Name) SELECT 2, SiteName from dbo.CUSTOM_SITES WHERE IsDeleted = 0
-
start all BigFix services
This procedure does not work for custom sites are not existing … in this case could try creating a new custom site with the same name and wait the time necessary the information reach the furthest agents in the deployment … at the end this custom site can be deleted …
Just as additional note, these procedures work in the hypothesis the agents are requesting for those operator sites or custom sites are reachable, otherwise it is another story.
As general consideration, in any case these warning messages are annoying but do not indicate the presence of problems.
2 Likes
Understood, and thank you for your insights we appreciate it.
So for custom sites that are not existing, 1.) recreate the custom site with the same name → 2.) Wait for endpoints to be subscribed again → 3.) Adjust the site's subscription criteria to no computers → 4.) Wait for a couple of days to make sure computers receive / evaluate the unsubscribscription, then finally delete the custom site?
yes, exactly, this is the idea.
One thing this product does not do well is clean up the mess it created.
Hi,
We have tried the procedure through iem CLI to propagate the opsites, but unfortunately, we still see gather errors for opsites on some of our relays based on latest relay log entries.
Sample errors from relaylogfile.txt
Date and time -0400 - 2380 - Unable to find ( or problems occurred ) with site id for URL: ``http://(RootFQDN):52311/cgi-bin/bfgather.exe/opsite1045`` - from (ENDPOINT IP)
Date and time -0400 - 2380 - 339: class NotASignedMessage - from (ENDPOINT IP)