Bug #17887
closedHosts with report listed for search query "not has last_report"
Added by El Joppa almost 8 years ago. Updated over 7 years ago.
Description
1.14RC2 fresh install.
Ubuntu Xenial.
Lists some hosts as with no reports even though it has reports:
Updated by Marek Hulán almost 8 years ago
- Subject changed from Hosts listed as with no last report to Hosts with report listed for search query "not has last_report"
- Category set to Reporting
The following query is being run for such search query
SELECT "hosts".* FROM "hosts" WHERE "hosts"."type" IN ('Host::Managed') AND ((NOT COALESCE("hosts"."last_report" IS NOT NULL, false))) ORDER BY "hosts"."name" ASC LIMIT 20 OFFSET 0
It seems to work just fine in my setup. Could you please check what's last_report for this host? You can do it in rails console with following commands
foreman-rake console p Host.find_by_name('ceph-008...').last_report exit
please upload the whole output here. Since you selected 1.14.0, does that mean you're hitting this issue with 1.14 RC2?
Updated by El Joppa almost 8 years ago
Yes, 1.14RC2 freshly installed on Ubuntu 16.04 with foreman-installer.
foreman console:
irb(main):001:0> p Host.find_by_name('ceph-008....').last_report Mon, 02 Jan 2017 09:26:44 UTC +00:00 => Mon, 02 Jan 2017 09:26:44 UTC +00:00 irb(main):002:0>
sql:
-[ RECORD 1 ]--------+-------------------------------- id | 100 name | ceph-008.X last_compile | last_report | updated_at | 2016-12-30 09:57:28.692717 created_at | 2016-12-30 09:57:18.963711 root_pass | architecture_id | 1 operatingsystem_id | 1 environment_id | 1 ptable_id | medium_id | build | f comment | disk | installed_at | model_id | 4 hostgroup_id | owner_id | owner_type | enabled | t puppet_ca_proxy_id | 1 managed | f use_image | image_file | uuid | compute_resource_id | puppet_proxy_id | 1 certname | ceph-008.X image_id | organization_id | location_id | type | Host::Managed otp | realm_id | compute_profile_id | provision_method | grub_pass | global_status | 0 lookup_value_matcher | fqdn=ceph-008.X pxe_loader |
Updated by Dominic Cleal almost 8 years ago
Are these config reports from Puppet in a default Foreman installation, or another config management plugin/tool? Perhaps last_report isn't updated if a different importer's in use.
Updated by El Joppa almost 8 years ago
All reports for all servers in foreman are from a seperate Puppetmaster with smartproxy and puppet reporter installed. Only some nodes has these lacking last_report.
Updated by Dominic Cleal almost 8 years ago
Could you tail -f /var/log/foreman/production.log
while running a Puppet agent on one of the affected hosts, and attach the resulting log? Perhaps an error is occurring during the report upload.
Updated by El Joppa almost 8 years ago
2017-01-03T11:22:56 ed625561 [app] [I] processing report for ceph-011. 2017-01-03T11:22:56 ed625561 [app] [D] Report: {"host"=>"ceph-011.", "reported_at"=>"2017-01-03 11:22:36 UTC", "status"=>{"applied"=>0, "restarted"=>0, "failed"=>0, "failed_restarts"=>0, "skipped"=>0, "pending"=>0}, "metrics"=>{"resources"=>{"changed"=>0, "failed"=>0, "failed_to_restart"=>0, "out_of_sync"=>0, "restarted"=>0, "scheduled"=>0, "skipped"=>0, "total"=>208}, "time"=>{"anchor"=>0.001184178, "apt_key"=>0.0018399150000000001, "augeas"=>3.808377495, "config_retrieval"=>7.508179066, "cron"=>0.000541659, "exec"=>0.010552525000000002, "file"=>0.28592155200000013, "file_line"=>0.000638734, "filebucket"=>0.000158476, "group"=>0.003276188, "ini_setting"=>0.007384523999999999, "ini_subsetting"=>0.000703383, "mailalias"=>0.0023759690000000003, "package"=>0.09707290800000001, "schedule"=>0.0009502489999999999, "service"=>0.244850981, "ssh_authorized_key"=>0.0032570940000000003, "sysctl"=>0.0015721229999999997, "total"=>11.989095714000001, "user"=>0.010258695000000002}, "changes"=>{"total"=>0}, "events"=>{"failure"=>0, "success"=>0, "total"=>0}}, "logs"=>[{"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Using configured environment 'production'"}, "level"=>"info"}}, {"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Retrieving pluginfacts"}, "level"=>"info"}}, {"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Retrieving plugin"}, "level"=>"info"}}, {"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Loading facts"}, "level"=>"info"}}, {"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Caching catalog for ceph-011."}, "level"=>"info"}}, {"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Applying configuration version '1483442566'"}, "level"=>"info"}}, {"log"=>{"sources"=>{"source"=>"Puppet"}, "messages"=>{"message"=>"Applied catalog in 5.52 seconds"}, "level"=>"notice"}}]} 2017-01-03T11:22:56 ed625561 [app] [D] Cache read: root_pass 2017-01-03T11:22:56 ed625561 [app] [D] Cache write: root_pass 2017-01-03T11:22:56 ed625561 [app] [D] Cache read: root_pass 2017-01-03T11:22:56 ed625561 [app] [D] Cache write: root_pass 2017-01-03T11:22:56 ed625561 [app] [I] Imported report for ceph-011. in 0.25 seconds
Is this enough?
Updated by Dominic Cleal almost 8 years ago
Yeah, it appears to be successful - the update should have happened before the last log message (though without SQL logs I can't tell if it did).
Regarding the earlier comment 2, is the record shown definitely the same as the host you looked up in the rake console command? The timestamps look off - the updated_at is 2016-12-30, but the last_report is 2017-01-02. It appears to me that you have multiple, perhaps differently named hosts.
Check with Host.find_by_name('ceph-008.X').id
or similar to check the record IDs are the same.
Updated by El Joppa almost 8 years ago
2016-12-30 is the day i installed Foreman and setup Puppet reports / facts push.
Actually they have different id's so these are just duplicates. Thats strange but i guess something has happened while setting up the puppet integration,
Updated by Dominic Cleal almost 8 years ago
- Status changed from New to Feedback
I'd suggest just deleting the older duplicate host showing up in the results, which should fix the issue.
I can't see why a duplicate would have been created, unless either the FQDN and/or domain differs between the two (unclear from obfuscation).
Updated by El Joppa almost 8 years ago
Deleted the duplicated failed hosts. They havent turned up yet.
Guess its just some glitch in the matrix while configuring Foreman/Puppet.