You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
New to Graylog.
I managed to get a cluster of 500 stateless CentOS 7 nodes installed with graylog-sidecar and filebeat. The feature I would like to have is a default sidecar collector when a new node joins, and persistent node-id:
Eg, when a node reboot, it will boot to the image, and start afresh, and it will rerun the graylog-sidecar -service install and get the new node-id. It will connect to the server, but with a new node-id, so I have to configure it to use the collector, it still shows two entries in the sidecar lists with new and old/inactive node-id, they should be merged. I have 500+ stateless clients, when they all reboot, they are all treated new(new node-id).
So two functions needed:
A: retain the node-id when a node-id file is polluted or lost,
B: ability to set a default sidecar collector config on the server, so nodes will work after reimaging and retained the logs.
I checked around briefly, a close request is issue #10454 but don’t find a track of work immediately. Apologies if I missed it. But it is a year from the previous issue.
Thanks,
Describe your environment:
OS Information: CentOS 7.9 3.10.0-1160.45.1.el7.x86_64
Service logs, configurations, and environment variables:
NA
3. What steps have you already taken to try and solve the problem?
NA
How can the community help?
A: retain the node-id when a node-id file is polluted or lost,
B: ability to set a default sidecar collector config on the server, so nodes will work after reimaging and retained the logs.
The text was updated successfully, but these errors were encountered:
Graylog 5.0 and Sidecar 1.3 will have support for tagged assignments again. #13433
Regarding your stateless setup, you can also define the node-id yourself directly in the sidecar.yml.
It does not have to be a file (which is the default).
The only importance is that it is unique. So you could even use your hostname, for instance.
Hi, Developers,
New to Graylog.
I managed to get a cluster of 500 stateless CentOS 7 nodes installed with graylog-sidecar and filebeat. The feature I would like to have is a default sidecar collector when a new node joins, and persistent node-id:
Eg, when a node reboot, it will boot to the image, and start afresh, and it will rerun the graylog-sidecar -service install and get the new node-id. It will connect to the server, but with a new node-id, so I have to configure it to use the collector, it still shows two entries in the sidecar lists with new and old/inactive node-id, they should be merged. I have 500+ stateless clients, when they all reboot, they are all treated new(new node-id).
So two functions needed:
A: retain the node-id when a node-id file is polluted or lost,
B: ability to set a default sidecar collector config on the server, so nodes will work after reimaging and retained the logs.
I checked around briefly, a close request is issue #10454 but don’t find a track of work immediately. Apologies if I missed it. But it is a year from the previous issue.
Thanks,
OS Information: CentOS 7.9 3.10.0-1160.45.1.el7.x86_64
Package Version:
graylog -4.2-repository-1-4.noarch
graylog -server-4.2.6-1.noarch
filebeat -8.0.0-1.x86_64
graylog -sidecar-1.1.0-1.x86_64
graylog -sidecar-repository-1-2.noarch
Service logs, configurations, and environment variables:
NA
3. What steps have you already taken to try and solve the problem?
NA
A: retain the node-id when a node-id file is polluted or lost,
B: ability to set a default sidecar collector config on the server, so nodes will work after reimaging and retained the logs.
The text was updated successfully, but these errors were encountered: