loading_request=0

[Deepak Singh]

I see almost all of my request logs has loading_request=0

We know that this is only present if the request causes the instance to be started, so what should i understand based on this? our app faces high latency most of the time.

Java, flexible env based app.

-- 

Tech

[Nick (Cloud Platform Support)]

Hey Deepak,

This means this was not the first request to the instance. This means that most of your requests are being handled by instances which already existed when the request was received by the App Engine front-end load balancer.

If you are still seeing latency, this could be caused by the instances being overloaded and the requests waiting in the pending queue in front of the instance (you can read about pending queues and scaling parameters in the app.yaml documentation (if using Java, see the appengine-web.xml documentation which is analogous)). In that case, you'll see another field, "pending_ms" in the log line. You can solve this type of latency by changing the scaling settings (see the docs).

If you don't see very high values for "pending_ms", check whether "cpu_ms" is comparable to the total request latency, as this could mean that your instance is taking a large amount of time to process the request.

If the "ms" field is high but the "cpu_ms" is not high, this could mean that a remote API call made during handling the request takes a long time to complete while the instance waits for it to return before sending the response.

Finally, if the "ms" field is not high compared to what the client observes, it's possible that the extra time is added during network transfer to and from the App Engine infrastructure and your instance. This could be solved by improving the network the client is on (not always a possibility) or locating the app closer to them (create an EU application if clients are in Europe).

You can read more about request log fields in the documentation.

I hope this short guide is helpful in determining the cause of the request latency. If you have any further questions, let me know and I'll be happy to assist!

Cheers,

Nick
Cloud Platform Community Support

[Deepak Singh]

Hi Nick,

Thanks for your detailed response. 

Based on your response, i guess that none of the data seems suspicious in our logs. Here is a log snippet

ms=214 cpu_ms=0 cpm_usd=0.00002771493 loading_request=0 instance=- app_engine_release=1.9.42 trace_id=c0ed6d9452c7fa778840af791c4a291d

However look at the below graphs where latency shows the data and cpu graph shows the enough instance. But still we have user latency.

Google page speed test indicates server response time as close to 0.5 sec which is quite high. 

CPU

% CPU

3 Aug, 12:003 Aug, 15:003 Aug, 18:003 Aug, 22:025101520

Latency

Milliseconds

3 Aug, 12:003 Aug, 15:003 Aug, 18:003 Aug, 21:5920,00040,00060,00080,000100,000120,000

 Average request latency: 

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.
To unsubscribe from this group and stop receiving emails from it, send an email to google-appengi...@googlegroups.com.
To post to this group, send email to google-a...@googlegroups.com.
Visit this group at https://groups.google.com/group/google-appengine.
To view this discussion on the web visit https://groups.google.com/d/msgid/google-appengine/301897c2-16ac-4d30-ad84-7f19e7a2d870%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Deepak Singh

[Deepak Singh]

Sorry, missed to write the data

Average request latency: 127.05

CPU: 8.996

No data for loading latency

--

Deepak Singh

[Nick (Cloud Platform Support)]

Hey Deepak,

Could you reupload those attached images? I can't see them from my end. As well, a few more example logs would help to determine what might be going on. 

Cheers,

Nick
Cloud Platform Community Support

On Wednesday, August 3, 2016 at 12:39:22 PM UTC-4, Deepak Singh wrote:

Hi Nick,

Thanks for your detailed response. 

Based on your response, i guess that none of the data seems suspicious in our logs. Here is a log snippet

ms=214 cpu_ms=0 cpm_usd=0.00002771493 loading_request=0 instance=- app_engine_release=1.9.42 trace_id=c0ed6d9452c7fa778840af791c4a291d

However look at the below graphs where latency shows the data and cpu graph shows the enough instance. But still we have user latency.

Google page speed test indicates server response time as close to 0.5 sec which is quite high. 

CPU

% CPU

3 Aug, 12:003 Aug, 15:003 Aug, 18:003 Aug, 22:025101520

Latency

Milliseconds

3 Aug, 12:003 Aug, 15:003 Aug, 18:003 Aug, 21:5920,00040,00060,00080,000100,000120,000

 Average request latency: 

On Wed, Aug 3, 2016 at 12:58 AM, 'Nick (Cloud Platform Support)' via Google App Engine <google-appengine@googlegroups.com> wrote:

Hey Deepak,

This means this was not the first request to the instance. This means that most of your requests are being handled by instances which already existed when the request was received by the App Engine front-end load balancer.

If you are still seeing latency, this could be caused by the instances being overloaded and the requests waiting in the pending queue in front of the instance (you can read about pending queues and scaling parameters in the app.yaml documentation (if using Java, see the appengine-web.xml documentation which is analogous)). In that case, you'll see another field, "pending_ms" in the log line. You can solve this type of latency by changing the scaling settings (see the docs).

If you don't see very high values for "pending_ms", check whether "cpu_ms" is comparable to the total request latency, as this could mean that your instance is taking a large amount of time to process the request.

If the "ms" field is high but the "cpu_ms" is not high, this could mean that a remote API call made during handling the request takes a long time to complete while the instance waits for it to return before sending the response.

Finally, if the "ms" field is not high compared to what the client observes, it's possible that the extra time is added during network transfer to and from the App Engine infrastructure and your instance. This could be solved by improving the network the client is on (not always a possibility) or locating the app closer to them (create an EU application if clients are in Europe).

You can read more about request log fields in the documentation.

I hope this short guide is helpful in determining the cause of the request latency. If you have any further questions, let me know and I'll be happy to assist!

Cheers,

Nick
Cloud Platform Community Support

On Friday, July 29, 2016 at 12:01:29 PM UTC-4, Deepak Singh wrote:

I see almost all of my request logs has loading_request=0

We know that this is only present if the request causes the instance to be started, so what should i understand based on this? our app faces high latency most of the time.

Java, flexible env based app.

-- 

Tech

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.

To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscribe@googlegroups.com.
To post to this group, send email to google-appengine@googlegroups.com.

Visit this group at https://groups.google.com/group/google-appengine.
To view this discussion on the web visit https://groups.google.com/d/msgid/google-appengine/301897c2-16ac-4d30-ad84-7f19e7a2d870%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Deepak Singh

[Nick (Cloud Platform Support)]

Hey Deepak,

Any update? I'm here and happy to assist. 

Cheers,

Nick
Cloud Platform Community Support

On Wednesday, August 3, 2016 at 12:43:33 PM UTC-4, Deepak Singh wrote:

Sorry, missed to write the data

Average request latency: 127.05

CPU: 8.996

No data for loading latency

On Wed, Aug 3, 2016 at 10:08 PM, Deepak Singh <deepaks...@gmail.com> wrote:

Hi Nick,

Thanks for your detailed response. 

Based on your response, i guess that none of the data seems suspicious in our logs. Here is a log snippet

ms=214 cpu_ms=0 cpm_usd=0.00002771493 loading_request=0 instance=- app_engine_release=1.9.42 trace_id=c0ed6d9452c7fa778840af791c4a291d

However look at the below graphs where latency shows the data and cpu graph shows the enough instance. But still we have user latency.

Google page speed test indicates server response time as close to 0.5 sec which is quite high. 

CPU

% CPU

3 Aug, 12:003 Aug, 15:003 Aug, 18:003 Aug, 22:025101520

Latency

Milliseconds

3 Aug, 12:003 Aug, 15:003 Aug, 18:003 Aug, 21:5920,00040,00060,00080,000100,000120,000

 Average request latency: 

On Wed, Aug 3, 2016 at 12:58 AM, 'Nick (Cloud Platform Support)' via Google App Engine <google-appengine@googlegroups.com> wrote:

Hey Deepak,

This means this was not the first request to the instance. This means that most of your requests are being handled by instances which already existed when the request was received by the App Engine front-end load balancer.

If you are still seeing latency, this could be caused by the instances being overloaded and the requests waiting in the pending queue in front of the instance (you can read about pending queues and scaling parameters in the app.yaml documentation (if using Java, see the appengine-web.xml documentation which is analogous)). In that case, you'll see another field, "pending_ms" in the log line. You can solve this type of latency by changing the scaling settings (see the docs).

If you don't see very high values for "pending_ms", check whether "cpu_ms" is comparable to the total request latency, as this could mean that your instance is taking a large amount of time to process the request.

If the "ms" field is high but the "cpu_ms" is not high, this could mean that a remote API call made during handling the request takes a long time to complete while the instance waits for it to return before sending the response.

Finally, if the "ms" field is not high compared to what the client observes, it's possible that the extra time is added during network transfer to and from the App Engine infrastructure and your instance. This could be solved by improving the network the client is on (not always a possibility) or locating the app closer to them (create an EU application if clients are in Europe).

You can read more about request log fields in the documentation.

I hope this short guide is helpful in determining the cause of the request latency. If you have any further questions, let me know and I'll be happy to assist!

Cheers,

Nick
Cloud Platform Community Support

On Friday, July 29, 2016 at 12:01:29 PM UTC-4, Deepak Singh wrote:

I see almost all of my request logs has loading_request=0

We know that this is only present if the request causes the instance to be started, so what should i understand based on this? our app faces high latency most of the time.

Java, flexible env based app.

-- 

Tech

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.

To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscribe@googlegroups.com.
To post to this group, send email to google-appengine@googlegroups.com.

Visit this group at https://groups.google.com/group/google-appengine.
To view this discussion on the web visit https://groups.google.com/d/msgid/google-appengine/301897c2-16ac-4d30-ad84-7f19e7a2d870%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Deepak Singh

--

Deepak Singh

[Deepak Singh]

Hi Nick,

I will soon update you as i am currently out of station.

To view this discussion on the web visit https://groups.google.com/d/msgid/google-appengine/f4c9e93b-0dc1-4c11-8c7b-af56f956f6a0%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Deepak Singh

[Deepak Singh]

Hi Nick,

Please find the attached graphs that will explain  the latency issues.

Regards
Tech team

On Aug 11, 2016 12:15 AM, "Deepak Singh" <deepaks...@gmail.com> wrote:

Hi Nick,

I will soon update you as i am currently out of station.

[Deepak Singh]

Hi Nick,

Any update.

Regards

Tech team

--

Deepak Singh