Background :
I was testing some custom scripts for graphing on my server cluster directly. Though working I would notice gaps in graphs and corresponding timeouts in execution. These were running locally on our proxy deployment but connecting to the server cluster DB to run the queries, and insert into custom DB for graphing. My thought was to move the same scripts to local proxy and query direct on the zabbix_proxy DB and allow the server<->proxy sync take care of the rest.
Issue:
I'm not sure if this is by design, an issue or expected behavior but there are empty columns in multiple tables all over the proxy DB that I know are being polled as the same columns on the server are there. My specific issue is items table with lld discovery items - example query
I was testing some custom scripts for graphing on my server cluster directly. Though working I would notice gaps in graphs and corresponding timeouts in execution. These were running locally on our proxy deployment but connecting to the server cluster DB to run the queries, and insert into custom DB for graphing. My thought was to move the same scripts to local proxy and query direct on the zabbix_proxy DB and allow the server<->proxy sync take care of the rest.
Issue:
I'm not sure if this is by design, an issue or expected behavior but there are empty columns in multiple tables all over the proxy DB that I know are being polled as the same columns on the server are there. My specific issue is items table with lld discovery items - example query
- SELECT key_ FROM items; *returns 24,000 rows with key_ and discovered value which should have a corresponding name in same row.
- SELECT name FROM items; *returns 24,000 blank rows?
- SELECT name FROM items WHERE NOT name=' ' OR NULL; *empty results, there is no name data being inserted into proxy DB
Comment