Error reporting and VM instance name

[Pavel Madr]

Hi,

We have a group of managed instances and they report errors via stackdriver error reporting.

Is it possible to see the name of VM which reported concrete error in Stackdriver Error Reporting web page? Or its IP address?

So we can connect to the specific instance and analyze the error more in its logs.

Best Regards,

Pavel

[Mary Koes]

Hi Pavel,

You should be able to click "View Logs" from the error detail page and drill into the specific logs from the VM that created the log entry.  

You'll also see related logs in context and can add or remove logs to consider.  It might be helpful from these logs to switch to the advanced logs viewer by clicking on the arrow at the far right of the filter menu in logging.

Best,

Mary
Product Manager, Stackdriver Error Reporting

--
© 2017 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 
Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-discussion@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.
---
You received this message because you are subscribed to the Google Groups "gce-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussion+unsubscribe@googlegroups.com.
To post to this group, send email to gce-discussion@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gce-discussion/d5aea124-3141-494a-b03b-9b5aacd5e961%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Pavel Madr]

Hi Mary,

Thank you for your answer. I'm trying logging and it works fine.

Best Regards,

Pavel

On Wednesday, October 11, 2017 at 8:30:22 PM UTC+2, Mary Koes wrote:

Hi Pavel,

You should be able to click "View Logs" from the error detail page and drill into the specific logs from the VM that created the log entry.  

You'll also see related logs in context and can add or remove logs to consider.  It might be helpful from these logs to switch to the advanced logs viewer by clicking on the arrow at the far right of the filter menu in logging.

Best,

Mary
Product Manager, Stackdriver Error Reporting

On Wed, Oct 11, 2017 at 2:55 AM, Pavel Madr <pavel...@gmail.com> wrote:

Hi,

We have a group of managed instances and they report errors via stackdriver error reporting.

Is it possible to see the name of VM which reported concrete error in Stackdriver Error Reporting web page? Or its IP address?

So we can connect to the specific instance and analyze the error more in its logs.

Best Regards,

Pavel

--
© 2017 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 

Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-dis...@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.

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

To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussio...@googlegroups.com.
To post to this group, send email to gce-dis...@googlegroups.com.

[Pavel Madr]

Hi Mary,

I want to use gcloud logging in Django too. It seems not to work in asynchronous mode (BackgroundThreadTransport).

If I set transport to SyncTransport, it works fine.

Can I use asynchronous gcloud logging in Django or Flask?

On Wednesday, October 11, 2017 at 8:30:22 PM UTC+2, Mary Koes wrote:

Hi Pavel,

You should be able to click "View Logs" from the error detail page and drill into the specific logs from the VM that created the log entry.  

You'll also see related logs in context and can add or remove logs to consider.  It might be helpful from these logs to switch to the advanced logs viewer by clicking on the arrow at the far right of the filter menu in logging.

Best,

Mary
Product Manager, Stackdriver Error Reporting

On Wed, Oct 11, 2017 at 2:55 AM, Pavel Madr <pavel...@gmail.com> wrote:

Hi,

We have a group of managed instances and they report errors via stackdriver error reporting.

Is it possible to see the name of VM which reported concrete error in Stackdriver Error Reporting web page? Or its IP address?

So we can connect to the specific instance and analyze the error more in its logs.

Best Regards,

Pavel

--
© 2017 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 

Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-dis...@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.

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

To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussio...@googlegroups.com.
To post to this group, send email to gce-dis...@googlegroups.com.

[Mary Koes]

Hi Pavel,

I checked with the client libraries team that builds the libraries for Python and they confirmed that asynchronous logging with the Python libraries should work in both Flask and Django.  Are you using the most up-to-date version of the client libraries?

Best,

Mary

To view this discussion on the web visit https://groups.google.com/d/msgid/gce-discussion/3a3575f8-cddd-498e-85ad-edbaaa2b842a%40googlegroups.com.

[Pavel Madr]

Hi Mary,

I'm sorry for the late response due to another issues.

I tried it again and the problem seems to be connected with using UWSGI and Django.

I'm using UWSGI option processes=2. If this option is not used, gcloud logging is working perfectly.

So it seems that there is a problem with logging when the process is forked (my Flask also runs via UWGSI). Also asynchronous logging doesn't work with RQ which does a fork too.

Please try:

$ uwsgi --module=mysite.wsgi:application --http :8000 --enable-threads --processes=2

and

$ uwsgi --module=mysite.wsgi:application --http :8000 --enable-threads

If you need more details, I can provide a sample Django project.

Best Regards,

Pavel

[Mary Koes]

Hi Pavel,

Just confirming that this can happen with older versions of the client libraries but wouldn't be expected with the current versions.  More specifically, there was an issue that could cause this with gRPC versions less than 1.7.

The next step in debugging the issue would be getting a full list of the libraries you have installed. (pip freeze can output the names and versions of all installed packages).

You could either open a ticket with support or else send me the info off list.

Best,

Mary

Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-discussion@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.

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

To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussion+unsubscribe@googlegroups.com.
To post to this group, send email to gce-discussion@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gce-discussion/bcd2d3b0-ef70-418e-bbb7-c874798b7971%40googlegroups.com.