Cognos Analytics

Hi,

After we upgraded to 12.0.2 around a month ago, some of the power users are reporting that they experience unstable performance, when running dashboards. This is both when creating and consuming.

Examples are expanding members in data source or in filter bar. Some times they appear immediately and other times they use 10-30 seconds before they appear. The same unstable performance is also seen when filters are changed for visualizations. eg. the query execution time is 0,5 seconds while the visualization some refreshes instantly while other times it takes forever, even the browser throws warning that the script is taking long time running.
The source is DMR-package

This doesn't seem to be related to the data, as I have investigated the queries, and they are quite predictable in their execution time, while Cognos is not. And it is seen both when the data sets are small and large.

I have also checked the dispatchers, and all of them seem to be running fine.

Are anyone seeing anything similar?

Hi Sandeep,

have you monitored the server resources during these waiting times?

We have a similar recurring issue that we have not yet been able to nail down to a distinct reason. It shows up randomly that the server CPU consumption goes up to 100% on all cores and stays that for about 30-60 seconds until the system responds again. Is that the same issue when you have to wait for the filter/viz to respond?

Hi,

After we upgraded to 12.0.2 around a month ago, some of the power users are reporting that they experience unstable performance, when running dashboards. This is both when creating and consuming.

Examples are expanding members in data source or in filter bar. Some times they appear immediately and other times they use 10-30 seconds before they appear. The same unstable performance is also seen when filters are changed for visualizations. eg. the query execution time is 0,5 seconds while the visualization some refreshes instantly while other times it takes forever, even the browser throws warning that the script is taking long time running.
The source is DMR-package

This doesn't seem to be related to the data, as I have investigated the queries, and they are quite predictable in their execution time, while Cognos is not. And it is seen both when the data sets are small and large.

I have also checked the dispatchers, and all of them seem to be running fine.

Are anyone seeing anything similar?

Hi Robert,

We are on Cognos Cloud hosted environment, so not able to monitor the servers our selves.

But great input, for a support case.

Are you also on 12.0.2?

Hi Sandeep,

have you monitored the server resources during these waiting times?

We have a similar recurring issue that we have not yet been able to nail down to a distinct reason. It shows up randomly that the server CPU consumption goes up to 100% on all cores and stays that for about 30-60 seconds until the system responds again. Is that the same issue when you have to wait for the filter/viz to respond?

Hi,

After we upgraded to 12.0.2 around a month ago, some of the power users are reporting that they experience unstable performance, when running dashboards. This is both when creating and consuming.

Examples are expanding members in data source or in filter bar. Some times they appear immediately and other times they use 10-30 seconds before they appear. The same unstable performance is also seen when filters are changed for visualizations. eg. the query execution time is 0,5 seconds while the visualization some refreshes instantly while other times it takes forever, even the browser throws warning that the script is taking long time running.
The source is DMR-package

This doesn't seem to be related to the data, as I have investigated the queries, and they are quite predictable in their execution time, while Cognos is not. And it is seen both when the data sets are small and large.

I have also checked the dispatchers, and all of them seem to be running fine.

Are anyone seeing anything similar?

Nope, it's a on-prem installation of V 11.1.7 FP7 but I just wanted to mention that we see that recurring spill up of CPUs from time to time. So far it seems like nothing to do with data but with metadata read from models.

Hi Robert,

We are on Cognos Cloud hosted environment, so not able to monitor the servers our selves.

But great input, for a support case.

Are you also on 12.0.2?

Hi Sandeep,

have you monitored the server resources during these waiting times?

We have a similar recurring issue that we have not yet been able to nail down to a distinct reason. It shows up randomly that the server CPU consumption goes up to 100% on all cores and stays that for about 30-60 seconds until the system responds again. Is that the same issue when you have to wait for the filter/viz to respond?

Hi,

After we upgraded to 12.0.2 around a month ago, some of the power users are reporting that they experience unstable performance, when running dashboards. This is both when creating and consuming.

Examples are expanding members in data source or in filter bar. Some times they appear immediately and other times they use 10-30 seconds before they appear. The same unstable performance is also seen when filters are changed for visualizations. eg. the query execution time is 0,5 seconds while the visualization some refreshes instantly while other times it takes forever, even the browser throws warning that the script is taking long time running.
The source is DMR-package

This doesn't seem to be related to the data, as I have investigated the queries, and they are quite predictable in their execution time, while Cognos is not. And it is seen both when the data sets are small and large.

I have also checked the dispatchers, and all of them seem to be running fine.

Are anyone seeing anything similar?

Thanks for the input, Robert.

I have created a support case. Lets see what they say

Nope, it's a on-prem installation of V 11.1.7 FP7 but I just wanted to mention that we see that recurring spill up of CPUs from time to time. So far it seems like nothing to do with data but with metadata read from models.

Hi Robert,

We are on Cognos Cloud hosted environment, so not able to monitor the servers our selves.

But great input, for a support case.

Are you also on 12.0.2?

Hi Sandeep,

have you monitored the server resources during these waiting times?

We have a similar recurring issue that we have not yet been able to nail down to a distinct reason. It shows up randomly that the server CPU consumption goes up to 100% on all cores and stays that for about 30-60 seconds until the system responds again. Is that the same issue when you have to wait for the filter/viz to respond?

Hi,

After we upgraded to 12.0.2 around a month ago, some of the power users are reporting that they experience unstable performance, when running dashboards. This is both when creating and consuming.

Examples are expanding members in data source or in filter bar. Some times they appear immediately and other times they use 10-30 seconds before they appear. The same unstable performance is also seen when filters are changed for visualizations. eg. the query execution time is 0,5 seconds while the visualization some refreshes instantly while other times it takes forever, even the browser throws warning that the script is taking long time running.
The source is DMR-package

This doesn't seem to be related to the data, as I have investigated the queries, and they are quite predictable in their execution time, while Cognos is not. And it is seen both when the data sets are small and large.

I have also checked the dispatchers, and all of them seem to be running fine.

Are anyone seeing anything similar?

Hi Sandeep, would you mind sharing what you find out? I'm sure it could help others!

Thanks for the input, Robert.

I have created a support case. Lets see what they say

Nope, it's a on-prem installation of V 11.1.7 FP7 but I just wanted to mention that we see that recurring spill up of CPUs from time to time. So far it seems like nothing to do with data but with metadata read from models.

Hi Robert,

We are on Cognos Cloud hosted environment, so not able to monitor the servers our selves.

But great input, for a support case.

Are you also on 12.0.2?

Hi Sandeep,

have you monitored the server resources during these waiting times?

We have a similar recurring issue that we have not yet been able to nail down to a distinct reason. It shows up randomly that the server CPU consumption goes up to 100% on all cores and stays that for about 30-60 seconds until the system responds again. Is that the same issue when you have to wait for the filter/viz to respond?

Hi,

After we upgraded to 12.0.2 around a month ago, some of the power users are reporting that they experience unstable performance, when running dashboards. This is both when creating and consuming.

Examples are expanding members in data source or in filter bar. Some times they appear immediately and other times they use 10-30 seconds before they appear. The same unstable performance is also seen when filters are changed for visualizations. eg. the query execution time is 0,5 seconds while the visualization some refreshes instantly while other times it takes forever, even the browser throws warning that the script is taking long time running.
The source is DMR-package

This doesn't seem to be related to the data, as I have investigated the queries, and they are quite predictable in their execution time, while Cognos is not. And it is seen both when the data sets are small and large.

I have also checked the dispatchers, and all of them seem to be running fine.

Are anyone seeing anything similar?