General web database question?

[Gasman]

Hi all,
We were moved by our telephone provider to their new platform.
One of the features we use is call recording and one of my tasks evey day is to download the previous days recordings.


This is accomplished, by logging into a site and selecting the relevant data of course.
Since moving to the new platform/system the performance has been abysmally slow. They are saying that is down to my computer as they can access it fine.?
I counter with

I had no problems using the same computer on the old system.
I have used another computer in the office with the same result

I have accessed it from home late in the evening (when the load should be low or non existent) and it is still slow.


Now as things are quiet, we are only talking about 600 calls for the week and 80-120 calls per day. the download is OK, just the presentation of the data for selection.
It occurred to me last night, that should my computer even have an effect.?
Isn't it a case of my selecting an option, the webpage sending the request to the server and the server sending the data back to the page.?


TIA

 

[jleach]

Sounds right to me... if the data is coming from their server, your computer shouldn't have anything to do with it. It's either a) their server is slow, b) you have reduced bandwidth for the download, or c) maybe something crazy like your computer is completed tapped out in resources, but I'd expect other things to start dying long before this.

Try using a tool like Fiddler or your browser's dev tools network profiler to see how long the download is taking (you might be able to tell from that whether the holdup is in bandwidth or in their server response time: if their response time is high, it's definitely on their end, which is the most likely case, I'd think).

 

[Gasman]

Hi jleach,
Downloading the recording as a zip file is not the problem, even if we are only on 40Mb download speed. It is the presentation of the call log data.
I have a filter set up for yesterdays calls and the page set to the max to be displatyed of 100.
However when I log in and select the recordings tab, it can take over 90 seconds to show the data BEFORE I can even get to apply my filter.?
Then I have to wait a similar time for perhaps 100 calls to appear.
For instance today, logging in it shows 2100 entries availble, and when filtered there are 129 entries to download.
I can only download a page at a time, hence the 100 per page.

This works the same as the old platform, only that ran miles faster.

In fact I compared both whilst I still had recordings in the old platform for the same period in days and in the old platform there would have been more calls for a week that we are doing now.

I've tried two other computers, one in work and my own persoanl laptop at home, all with the same result.?

We use other web based databases as well of course and have no problems with those.
This has only occurred since moving to the new platform.

So I was asking if my thinking is correct, in the way that the webpage and the database interact?

 

[jleach]

Sorry, not quite following.

You're viewing their website running off their data and it's really slow?

Or you have an Access app that's connected to a remote data set that they host?

If it's the former, I can't see how it could possibly be an issue on your end. The latter, maybe, but I'll hold off until I know for sure what's going on.

Cheers,

 

[Gasman]

No nothing to do with Access, just a general DB question.
The former is the situation, yet they tell me all is fine at their end.

Just wanted input from more knowledgeable people.

Personally I think it is BS, and whilst I can be gullible at times, I can also become a little sceptic

 

[The_Doc_Man]

When you have the same problem with three different computers, that should tell you that it isn't your system. I take it from your description that you are somehow accessing a web page and choosing options to download a file.

You need to know that there is such a thing as a web activities log that includes a timestamp of every page access request and every download action. However, how you enable logging differs for every browser and you would have to look up that detail for your specific browser.

If you GET the logs, they are a big, ugly, linear file with various data elements that, at least for the first four or five fields, are predictable in order and general format, but NOT predictable in spacing. It is a case where you might see fields like those below (but the exact order depends on the browser):

* Sequence number (monotonically increasing integer, usually)

* Timestamp logged at least accurately enough to show time to the second (but sometimes to the millisecond depending on the browser). AND TCP/IP will DEFINITELY use the millisecond timer at least internally, so that info IS available.

* An HTTP transaction name, always one word, may include GET or PUT but many other keywords exist

* Some sort of web-site designation, http:\\ or https:\\ type of reference

* Other data of varying formats.

If you can't enable logging with your browser, check system logs using Start >> Control Panel >> Administrative Tools >> Event Browser to see if anything is being logged while you have that page open.

NOTE: Sometimes, enabling logging requires you to find the browser's home directory (where all of its configuration files are stored) and use Notepad to edit a .INI or .CFG or .DAT file to change logging from level 0 to something higher than 0. IF that is what it takes, make a copy of the affected config file before editing and when you are done with the logging, RESTORE that config file without editing. In that case, you will need to restart the browser after changing the config file, because config changes are launch-time events.

 

[Gasman]

Thank you The _Doc_Man,
However I cannot make head or tail of that log.
I am using Chrome as my main browser. However I also use Firefox and Pale Moon.
The issue is in all browsers.

I think you have confirmed what I thought was the case at the start.
The weird thing is all the tasks that I would have thought were more cpu intensive, like download and exporting, work reasonable quickly

 

[The_Doc_Man]

Remember that the log of an HTTP / HTTPS file is transaction oriented. The browser asks for something and the remote host provides it. The time tags would indicate stuff requested by YOU and also by your browser based on embedded pictures and other content. By comparing times you can develop a sense of how long each page is taking and what is on it.

Most of the browser logs I have seen are space-delimited single-line logs in Notepad readable format (and the lines can get QUITE long for ugly web site architectures). You have to look at them to find the place where YOU requested something and then note a cascade of events where embedded content comes down. Frequently, but certainly subject to the whims of the page designer, embedded content like pics and graphics will come last.

The shame is that your backups probably DON'T have copies of older logs in which you could compare load times and PROVE your point. Because slow loading of complex pages is strictly on the shoulders of the host's designers and managers.