I've recently been in the situation where I have introduced monitoring onto SQL databases and onto an environment including a SAN. The customer has told me firstly that the alerts being generated are too noisy, then secondly that we should stop access to individuals and stop monitoring all together on some databases.
Could I reduce the number of alerts?
I have gone back to the customer and pointed out that the alerts are valid and I have checked through the thresholds with the customer and we have confirmed that they are valid thresholds.
They were receiving emails every day to tell them about problems with databases and nodes.
Database access denied.
The SQL server databases that I was monitoring suddenly stopped being monitored and then the DBA's came and told me that the Solarwinds account no longer had access due to excessive SQL reads on the cluster and this was causing a problem for the databases.
One of the project managers then came to me, to ask me why his report on SQL databases had stopped working?
Embarrassed technical engineers.
The SAN was recently monitored too, with SRM, it highlighted numerous issues as soon as it was put in, all the time during meetings the technical engineers were assuring the managers and directors that the SAN was fit for purpose and there was plenty of space.
One of the managers who had Solarwinds access logged in to Solarwinds and saw the state of the SAN and the problems it had connecting to a SQL database that had particular latency issues. After the meeting I was told to stop access to SRM for this particular manager.
Solarwinds is great and information I gather about our infrastructure is invaluable. Sometimes Solarwinds can be its own worst enemy.
I am discovering that I need to be a politician as well as a monitoring guy to keep the monitoring alive and to give people the information they need. Different people within the organisation have different requirements and I have to juggle their needs.
Has anyone else experienced this?