This is just a small bug-fix release of the plugin. It has actually been quietly released for a while now. If you have downloaded the plugin recently, you have the latest version. To be sure, check the version in the Console, or you can see it in the file name.
There are two bugs fixed:
1. Deployment on an OMS hosted on Solaris didn’t work. (And I suspect it could be the same for Agents on Solaris.)
2. Changing thresholds on the metrics caused the error “Modification of Target Monitoring Settings has Failed”. Also, applying monitoring template was failing for the same reason.
If you didn’t know that you could setup threshold and customized the thresholds that are set out of the box with the plugin, then you are not fully using Enterprise Manager. You can do that in the MySQL target menu — Monitoring -> Metric and Collection Settings. When alert or warning conditions are met, Oracle EM can raise incidents and send notification according to your settings. As I stated above, 12.1.0.1.0 had a bug that prevented users from changing the thresholds.
Same known issues exist as for 12.1.0.1.0 release.
Useful links:
Follow the same deployment procedure for upgrade — (1) import into EM12c using EMCLI, (2) upgrade on OMS by deploying on OMS, (3) upgrade on Agents byt deploying on agents.
Happy monitoring!
21 Comments. Leave new
Is is “Quite release” or “Quiet release”?
Or maybe it was not really quite a quiet release? :-)
Hehe… just checking if anybody is paying attention. :)
This is excellent, Alex. Fixed the metric and template issues I has having with the previous version. And works with both OEM 12.1.0.1 and 12.1.0.2. Nice job!
Hi Alex,
we are using your MySQL-Plugin 12.1.0.1.2 with Cloud Control 12.1.0.2. and are really excited.
Sadly, there is one problem. The monitored MySQL databases logs a warning in the error log every five minutes:
[Warning] Aborted connection 7011 to db: ‘unconnected’ user: ‘our_user’ host: ‘our_host’ (Got an error reading communication packets)
Perhaps, is there a missing close_mysql_connection or something else?
Regards,
Spiker
Hi Alex,
solved it. I omitted the REPLICATION CLIENT right. We doesn’t use replication, so i thought this user right isn’t necessary.
Regards,
Spiker
Interesting… Thanks for reporting. I’ll check to see if there is a better way to handle it in the plugin than leaving unfinished connection.
Hi Alex,
I am getting 404 file not found, when trying to download the plugin. Also, can you confirm if I need to buy any licenses from oracle to use these plugins. Just to give a bit of background, I am currently using cloud control 12.1.0.2 to monitor my oracle databases. I am using a base installation of OEM with diagnostic and tuning packs enabled for oracle.
The Mysql databases that I am planning to add are all community editions and I do not have any support for the same.
Regards,
annuj
Anuj: I’ll ask out web team to check download issue. Must have missed links update on web-site relaunch which we’ve just completed.
Regarding the costs — any cost to run EM12c and monitor any targets that Oracle provides is your responsibility. The free version of the MySQL plugin itself downloaded from Pythian web-site does not have licensing fees. However, any other targets you monitor is a different story including monitoring the host itself where you are running the MySQL and local agent if you monitor MySQL by deploying the client on the host.
Commercial support for MySQL plugin is available too if you are interested — just fill Contact Us form.
Hi
The links are working now, thank you.
Hi Alex,
We are using cloud control version 12.1.0.3 . To add MySQL server we only see the option to add MySQL Instance and not the MySQL Database.
How can we add the MySQL database also.
Regards,
Nishant
Nishant: I used term instance to avoid confusion with term database (that is basically a schema in MySQL). You don’t need to add anything else.
Hi Alex,
We have CC 12104 and we have recently added the 12.1.0.1.2 MySQL plug-in and started to monitor some My SQL databases. When we look into the Incident Manager in Cloud Control, we see that the My SQL incidents are the only ones that don`t have a Category assigned and our incident and notification rules are based on that. Do you know why the My SQL incidents don`t have a category associated with an incident?
Cheers,
Florin
Hey Alex,
So I’m trying to deploy the MySQL plugin into 12cR3. I get the plugin imported to the library and when I get ready to Deploy to the OMS Server it keeps failing with “java.sql.SQLException: ORA-01017: invalid username/password; logon denied.” I’ve verified that I’m giving it the right SYS password for my EM repository and I’m able to log in to the repository as sys from sqlplus. Any thoughts on what the real problem could be.
Thanks,
Sean
Hi Alex,
we are using the plugin to monitor a couple of mysql servers. After upgrading to 5.7. the plugin doesn’t work anymore for databases using the charset utf8. The query >select “alive”;select “alive”;
ERROR 1054 (42S22): Unknown error 1054
The double quotes seem to be the problem. with single quotes the query executes fine:
DTTAWST [email protected] [(none)]>select ‘alive’;
+——-+
| alive |
+——-+
| alive |
+——-+
1 row in set (0,01 sec)
The problem only occurs on databases with utf8, latin1 for example works fine.
Is this a known issue?
cheers
Sebastian
Hi Alex,
My colleague Sebastian Gentil mentioned in a comment before, that we have a problem with the plugin as we run into errors like:
ERROR 1054 (42S22): Unknown error 1054
In the meantime we have a more precise cause of the error. The problem ist that sql_mode ‘ANSI_QUOTES’ is used (required by the application). But with this sql_mode set, the plugin-check-scripts seem to run not properly.
Is there a way to work around that? The sql_mode ‘ANSI_QUOTES’ is required.
Thanks and best regards
Marcus
Hi Marcus,
Well, I guess a simple fix would be to replace the double quotes to single quotes.
However, I stopped actively maintaining the plugin since Oracle released their official one. I would need resurrect the build environment, do the change, test and re-package. I simply don’t have time for it at the moment to do pro-bono.
Alex
Hi Alex,
thanks for the reply. I didn’t know that your pluging is no longer maintained. It seems I found a way to work around the Problem by Setting a different sql_mode just for the monitoring user.
Best regards
Marcus
You can try to hack around that by finding this in the files in jar archive and (that’s what .opar is) and replace there. Though… without new version of opar upgrading it won’t work so it’s removal and installing new plugin.
Hi Alex,
hope you can help me on this one …
is it possile to change the retention time of the data collected by your plug-in?
Detailed data is only available for the last 24 hours, after that I only see one value per hour.
For example for the “slave lag” metric which is collected every 5 minutes.
I would like to keep all the detailed data for the last 30 Days.
thanks
Sebastian
Hi Sebastian,
This is not controlled by the plugin – it’s OEM settings. Here is some good starting point for you – https://dbasolved.com/2014/01/11/changing-oem12c-repository-retention-settings/
Hello, Alex.
I really need pythian.mysql.prod.oms.plugin_12.1.0.1.2 plugin, but here I can download only 12.1.0.3.0 version. Could you please share 12.1.0.2.0 version download link?
Many thanks and good luck!