Stuart D (Customer) asked a question.
Hi all,
My base understanding is that if we have 10 switches and we want to send their syslog data to WuG Total Plus, then each of those 10 nodes would consume 10 points.
But the bit I want to know is does anyone have insight on the impact on the server in terms of compute and storage requirements?
Currently we only run monitoring/alerting from WuG TP and its DB is a full version of SQL 2019 (IIRC) that runs on the same server as WuG. But, we would be looking to enable approximately 500 nodes to send their syslog data to WuG. If we did this, then I suspect our current specs won't cut the mustard:
Would the base advice be to have a separate DB Server for 500 devices syslogging to it and what minimum specs are we looking at?
WhatsUp Gold itself is NOT a syslog server. The SysLog Passive monitor, you SHOULD NOT use the "ANY SYSLOG" passive monitor. The Syslog passive monitor is designed to only accept syslog messages, those 1053 byte chucks which MATCH what you have in the syslog passive monitor.
WhatsUp Gold does have Log Management. Log Management does have a syslog server component of it which you can send all your syslogs to it. However, Log Management uses a remote Elasticsearch server db to store the Syslogs on that server and NOT in the WhatsUp Gold database where the Syslog Passive monitors are stored. Also you cannot run the Syslog Passive Monitor along with the Log Management or visa versa.
As the System requirements state, "the Elasticsearch SHOULD NOT be installed on the same machine as WhatsUp Gold". HERE are the specs for Elasticsearch
Thanks Mike.
> WhatsUp Gold itself is NOT a syslog server
This much I had worked out 😁
> the Elasticsearch SHOULD NOT be installed on the same machine as WhatsUp Gold".
Understood, and perhaps the most vital takeaway of your reply.
Muchas gracias.
Here's the KB on installing Elasticsearch for WUG and here is the latest Elasticsearch which will work.