Hi community, Hope all u r well.
I'm creating custom checkpoint templates on sqlserver probe, specifically one for obtain the ones that consume more cpu resources (called "heavy_queries"), I have the sql query for that and I tested it on the query tab inside edit template window with success.
The result below is from the query, I put the column "text" as row key and used the AvgCPUTime as the metric but for some reason the new QoS dont appear on the UMP. I created other template but this one does not have a long string text and looks fine and appear this QoS on the UMP.
My question is is not possible use long string text as row key?
Same issue occurs with the native sqlserver metric long_queries that have the column "sql_text" same as the "heavy_queries". The QoS long_queries appear on the UMP but does not show anything.
This is the query result of long_queries of sqlserver probe.
I'll appreciate your comments.
Did you check with Dr. Nimbus if something is posted on the QOS queue? It might be the probe doesn't send it? Is there anything in the log (with loglevel 5)?
Hi Christophe, thanks for your feedback!
No, I do not check yet with Dr. Nimbus, I didn't know about that tool since I'm a beginner with this platform
I opened it and it looks like very useful. I'm gonna perform some tests again and watch if there is something that can help to figure out this, also I'll change the loglovel to 5 in order to obtain more information about the probe.
I will post if I found something and if not too.
Hi @Emilio Ogaz Did you do manage to do any progress with your problem?
Hi Marco, yes I just answered about the last action performed and what I did to handle the problem.
Hi Christophe, Hope you are well.
I used Dr.Nimbus and I dont see any error or warning on the probe according this custome checkpoint. Also I put the loglevel on 5 to see more details and I noticed on the gathering information part of the checkpoint that the sql_string is too long (aren't short queries)
Finally I decided to use other column contained on the Query as a row key (login_name) and cpu_time as the metric.
Have a nice weekend!