[mapguide-users] Memory leak with MGOS 4 Beta 1

[Jackie Ng]

You could activate web tier logging of requests in the webconfig.ini which
should capture execution of mapagent operations like
GETDYNAMICMAPOVERLAYIMAGE that would happen on the web tier, but not the
server tier.

QueryMapFeatures on the server tier is merely servicing the selection/hover
queries driven by tooltips or map click/box selection from the map viewer.

On the web tier QUERYMAPFEATURES does more than QueryMapFeatures call to
the server tier, The mapagent operation has an assortment of parameters
that control how the base QueryMapFeatures server tier response is to be
composed into the final XML response to send back to the client (whether
you want to include feature attributes of selection, an inline selection
image, etc, etc)

- Jackie

On Thu, 8 Feb 2024 at 00:57, David Bowen <dbowenrci at gmail.com> wrote:

> Thanks Jackie. Is there anything else that we could be looking at to
> determine where the leak could be occurring? Is there any way to log which
> layers are turned on in the map?
>
> Also, what is the QUERYMAPFEATURES entry in the access.log doing? We've
> noticed that for an address searched, there's a random number of polygons
> drawn around the address but none seem to correlate with the geocoded
> coordinates.
>
> Image 1: The QUERYMAPFEATURES (red polygons) in relation to the address
> searched (blue dot).
> [image: image.png]
>
> Thanks,
> David
>
> On Sat, Feb 3, 2024 at 3:23 AM Jackie Ng <jumpinjackie at gmail.com> wrote:
>
>> Ok failing that, I'll pencil in a memory leak sweep in the beta 2 dev
>> cycle for the King Oracle provider just like I did previously for the GDAL
>> provider.
>>
>> https://trac.osgeo.org/fdo/ticket/1005
>>
>> - Jackie
>>
>> On Sat, 3 Feb 2024 at 06:12, David Bowen <dbowenrci at gmail.com> wrote:
>>
>>> We've looked through the access.log but nothing stands out in the
>>> preceding activity. There's nothing specific during the memory ramp up
>>> period that we don't see at other times.
>>>
>>> David
>>>
>>> On Wed, Jan 31, 2024 at 12:01 AM Jackie Ng via mapguide-users <
>>> mapguide-users at lists.osgeo.org> wrote:
>>>
>>>> Could you correlate the time period of 11.15am - 11.45am (where the
>>>> memory usage ramps up) with entries in your access.log and see which
>>>> operations are involved?
>>>>
>>>> - Jackie
>>>>
>>>> You wrote:
>>>>
>>>> Hello Pierre,
>>>>
>>>> The server is running Windows 2022 Standard, we're using Apache Server and
>>>> Oracle database. There aren't any error messages generated during the
>>>> memory leak. We've monitored the Memory usage on this server and notice
>>>> that memory consumption increases during the incidents (image 1). We tried
>>>> turning on/off layers, while monitoring the memory usage, which is how we
>>>> figured out the CLOBs issue but we haven't been able to figure out what's
>>>> causing the remaining memory leak.
>>>>
>>>>
>>>> Image 1: percent memory utilization on the server showing the increase
>>>> until the MapGuide services become unresponsive at the peak and have to be
>>>> restarted.
>>>> [image: image.png]
>>>>
>>>> _______________________________________________
>>>> mapguide-users mailing list
>>>> mapguide-users at lists.osgeo.org
>>>> https://lists.osgeo.org/mailman/listinfo/mapguide-users
>>>>
>>>
>>
>> --
>> http://themapguyde.blogspot.com
>>
>

-- 
http://themapguyde.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapguide-users/attachments/20240208/79233b4a/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 164763 bytes
Desc: not available
URL: <http://lists.osgeo.org/pipermail/mapguide-users/attachments/20240208/79233b4a/attachment.png>