mandevil · Product Manager · ✭✭✭✭✭

I don't, but if you open a support ticket, I'll bet they could walk you through it in 5 minutes. We have general alert creation documents, but not specific to any one alert...

OK, or if you can do a call today, just email support@confio.com (that email will only be good until the end of Sept. but still works now). If you do, we can send the GTM info...

Yes, this can be set up using any kind of logic or schedule (I won't get into those specifics as there are MANY ways one could implement this). What I meant is that the command and status *should* be consistent. If the command is set to 'START' and status is 'STOPPED', likely something has gone wrong and I'd look in the…

Yeah, this is beyond product support - more consulting at this point. 8 ) Unfortunately, I don't really free up until after the turkey day holidays. Want to try to connect after that?

Try a query against the CON_EVENT_SUM_<id> table. You can pretty it up, but something along these lines. This would be to get daily/hourly wait times by wait type. Again, haven't tested, but let me know if this gets you close to what you are looking for. set @instanceName := '<instance_name>'; set @dbid := (SELECT ID FROM…

I kind of like option 3. Seems like the algorithm for option 1 may prompt more questions and might lead me to believe the other represents all other wait time for that time period rather than just something that had been observed in a prior period but is not in the top X during this one. This method would likely drive…

I was thinking something like this then: select datediff(mi,(select max(date_created) from table)),getdate()) and use that value to compare with thresholds since it will be a numeric. So set warning between 15 and 30 minutes and critical over 30 for example. Set the alert to run at 15 minute intervals to keep overhead low.

How many records do you have in the interfaces table?

Manual Install Steps: ===================== 1. Login to Windows with an Admin account. 2. Unzip the DPA installation zip file to the desired the Program Files directory on a local drive, for example: c:\program files Note: If Windows User Account Control is on, you may need to create a directory somewhere else, unzip to…

I would agree from an overall perspective. Also, time spent in memory/cpu is what's ideal (could still be going after too much data due to missing indexes or inefficient queries). Time in memory/cpu means you are finding the data you are going for in cache. Basically means your processing and not in a suspended state.

There's a way to augment the deadlock alert. Since we can parse the error log, (make sure you set your trace flags appropriately) you can also configure an alert to send you some of the text from the error log. It won't be the complete stack, will point you right to where to find it in the error log.

Preaching to the choir. Has something to do with how things are tracked from an accounting standpoint per my understanding (think SKUs). They *should* all be in sync though.

Yeah, doesn't clean out old data. Either re-define the report to a shorter time period so that hash isn't included or wait until it ages out. Another option would be to define the report with custom SQL instead of the top N and include all hashes except the one you want to exclude.

Mark, I love it when I'm right! 8 ) Glad it helped you in figuring out what's going on. Reach out to us with anything else, you know where to find us. Rob

Azure shouldn't matter. I can think of a couple things that may be contributing, but this may be easier to cover on a quick call. Do you mind direct messaging me with your email and I'll set up a meeting that hopefully works time-wise.

Can you send your logs to support@confio.com? That option is still available for a few more weeks before it's to be discontinued. It will automatically open a ticket in our system also. To get the logs, go to options -> support tab -> zip all logs.

No update to the quick poll for specific collection. You'll have to still use the workaround.

Can you send me a table description for the one out there on RDS + a description on the one that's working? There should be a default value in the CONTT table for PUB that I suspect is not there in RDS...

What version of DPA are you running? Wonder if the schema doesn't match the version (as in a new column)? Mine are all populated with a 'Y'. Let me dig a bit more.

Basically, each target entered into DPA to be monitored requires a license. If this is to troubleshoot performance issues, you can submit a support case and they likely can kick you a temp license while troubleshooting.

might be easier to show you than describe it here. Would you be up for a screenshare?

Ha! None from me. The world runs on legacy.

I did a google search and came up with a lot of potential options. I'd rather stay agnostic at this point. 8 ) Some are free, others are at a cost for license. Looks like most are compatible with Oracle RDBMSs though. I have not tried any of them at this point though, so could not really venture pros and cons. Has anyone…

The feature request has not made it into the product at this point. It has some votes, but certainly not at the top of the list (top billing requests are much more likely to get actioned). What version of DPA are you at? There may be a way to accomplish this using DPAIM (the Orion integration module)...

You are most welcome. Here is how we go after the raw data (we grab the offsets and sql handle during our quick poll, then go back much less frequently to get the stats. SELECT sql_handle, statement_start_offset, statement_end_offset, plan_generation_num, plan_handle, SUM(execution_count) sum_execution_count,…

Yep, I see your point. If the status hasn't changed, then we probably shouldn't keep sending notification. You may want to open a support case as we may need to collect logs and potentially even go after raw data written to our alert tables in the repo.

Can you open a support case? There are specific options that you have to enable to get the blocking information. We'll want to make sure everything is enabled for your monitoring...

So that advice is coming from SQL Server directly, so I wonder why it's not capturing the db name? I take it CLA is the schema. Plan is used for multiple databases? Likely. OK, let's try this. We can join back to CONSW and then on to CONO. Forgive formatting, Oracle guy... Should work in SQL though. select db.NAME from…

Yep, I realize 2 issues, was just hoping to get the proverbial 2 birds with 1 GTM. 8 )

The blocking data is built from the detailed active session poll (we call it the quick poll). You can't keep more blocking data without keeping all of the detailed history. Having said that, yes, you can go to Options -> Administration Tab -> Advanced Options -> DB Instance Options tab -> make sure support options is…