NetXMS Support Forum

Migration to TimescaleDB solved this problem. At least it doesn't occur now  .

At first I thought that the problem was caused by duplicate data (see https://www.netxms.org/forum/general-support/migration-from-postgresql-to-timescaledb-failed/), but even after deleting duplicate data, the problem occurs with the old PostgreSQL database, but not with data migrated to TimescaleDB.

I think deleting the duplicate records will solve the problem.
Can I do it or will you need same more tests on my DB?

It returns 2 rows.

The table idata_160 has no primary key. Better to say that none of the idata_??? tables has a primary key. It is defined only on the idata table.
The idata_??? tables have item_id and idata_timestamp index only.

I've tried to migrate the data form PostgreSQL format to TimescaleDB. But it failed with following error messages:


# nxdbmgr migrate /etc/netxmsd.conf.pgsql
NetXMS Database Manager Version 3.3.340 Build 3.3-340-gdde893b7a5 (UNICODE)

Source:
        Driver: pgsql.ddr
        DB Name: netxms
        DB Server: 127.0.0.1
        DB Login: netxms

Target:
        Driver: pgsql.ddr
        DB Name: timescale
        DB Server: 127.0.0.1
        DB Login: netxms

Options:
        Skip audit log.............: no
        Skip alarm log.............: no
        Skip event log.............: no
        Skip syslog................: no
        Skip SNMP trap log.........: no
        Skip collected data........: no
        Skip data collection schema: no

Confirm database migration? (Yes/No) y
Database driver pgsql.ddr loaded
Connected to source database


WARNING!!!
This operation will clear all configuration and collected data from destination database before migration.
Are you sure? (Yes/No) y
Database successfully cleared
Migrating table config
Migrating table config_clob
Migrating table config_values
Migrating table users

...

Migrating table idata_160 to idata
Migrating table tdata_160 to tdata
Migrating table idata_164 to idata
ERROR: unable to insert data to destination table (23505 ERROR:  duplicate key value violates unique constraint "2161_2161_idata_sc_default_pkey"
DETAIL:  Key (item_id, idata_timestamp)=(1227, 2020-05-19 02:01:11+02) already exists.)
ERROR: unable to insert data to destination table (23505 ERROR:  duplicate key value violates unique constraint "2161_2161_idata_sc_default_pkey"
DETAIL:  Key (item_id, idata_timestamp)=(1227, 2020-05-19 14:25:13+02) already exists.)
ERROR: unable to insert data to destination table (23505 ERROR:  duplicate key value violates unique constraint "2161_2161_idata_sc_default_pkey"
DETAIL:  Key (item_id, idata_timestamp)=(1227, 2020-05-19 05:13:12+02) already exists.)

...

ERROR: unable to insert data to destination table (23505 ERROR:  duplicate key value violates unique constraint "2721_2721_idata_sc_default_pkey"
DETAIL:  Key (item_id, idata_timestamp)=(1227, 2020-05-28 22:45:14+02) already exists.)
ERROR: unable to insert data to destination table (23505 ERROR:  duplicate key value violates unique constraint "2721_2721_idata_sc_default_pkey"
DETAIL:  Key (item_id, idata_timestamp)=(1227, 2020-05-29 01:59:14+02) already exists.)
Database migration failed.

The server version is 3.3.340 on Debian.
Did I do something wrong or PostgreSQL uses different keys than TimescaleDB?
How can I correct the problem?

I wanted to configure a common graph for several DCI values with the same instance (they have same Instance Discovery script).

I used {instance-name} in the Title and Group fields, but it didn't work. All instances were displayed in one graph and the displayed Title of the graph contained the text {instance-name} instead of its value.
Then I tried to use {instance} in the Title and Group fields and everything works as I need. Even the title of the chart contains the name of the instance, so the contents of {instance-name}, even though I have {instance} used in the Title configuration.

it seems confusing to me and is probably not an intention.

The video uses the same URL format I described above. The URL connects to vCenter server, but the connection is restricted to a single ESXi server.

Following is written in the libvirt documentation https://libvirt.org/drvesx.html
"A vpx:// connection is currently restricted to a single ESX server. This might be relaxed in the future. The path part of the URI is used to specify the datacenter and the ESX server in it. If the ESX server is part of a cluster then the cluster has to be specified too."

This means that if I list all VMs I see only the VMs on this single ESXi server. If a running VM is moved to another ESXi server in the same cluster it disappears form the list and appears in the list for  the other ESXi server which needs another connection to the same vCenter but with different path. I'm not sure it will be possible to identify, that it is the same VM and move the DCIs instead of deleting old and creating new ones.

I know than this is property of livirt and it must be changed there but it limits the VMware monitoring in NetXMS.

This works fine:
virsh -c vpx://[email protected]/Datacenter_name/Cluster_name/ESXi_Server_name?no_verify=1

But it means I need one connection for each ESXi server and not one common connection for whole Cluster or for whole vCenter.

The answer is as I expected.

error: failed to connect to the hypervisor
error: internal error: Expecting 'vcenter.mydomain.local' to be a ESX(i) host but found a vCenter/VPX host

Thank you for your help.

Server: 3.3.323 on Debian 10

Web MC: 3.3.314 on Tomcat 9, Debian 10
Web browsers: Firefox, Edge, Chrome all on Windows 10

Windows MC: 3.3.314, Windows 10

One remark.
I've tried to replicate the problem with a template using Dummy values, but the problem never occurred with this template. It is very strange.

Try "host/nxmc-3.3.314".
I mean without / at the end.

I've tried to use vmgr.nsm. It works fine with ESXi hypervisors but when try to connect to a vCenter server it fails with following error:
2020.05.21 11:11:47.611 *W* [                   ] VMGR: Unable to add connection URL "esx://[email protected]/?no_verify=1" for connection name "vcenter"

The configuration is following:
[vmgr:vcenter]
Url = esx://[email protected]/?no_verify=1
Login = root
Password = XXX


The password in the real configuration is different but it is OK. I can login with this login and password to vCenter Web UI.

Did I do something wrong or supports vmgr.nsm ESXi hypervisors only?

Sometimes is the size of the graph not correct and it overlaps the adjacent graph.

I've a stacked graph with 7 values on the Performance tab and it is sometimes presented not correctly.
When I change the graph to not stacked it is always correct.

I thing it is similar problem I reported for the executeSSHCommand few weeks ago: https://www.netxms.org/forum/general-support/executesshcommand-reports-getlistvalue()-result-is-500-(internal_error)/

Thank you.

Similar problem is with some other parameters of the the graph. For example Time period and Y Axis Range, but this is not so confusing.