Padanfain wrote:
A curiosity... Will the single toon submission from a cross realm sighting give
an update toward the 100 updates per 30 days needed to make that realm
census 'reliable data?'
that is a Rollie question but from what I've been told that might just be the case.
the site doesn't have any way to tell if you did full census runs or just started and killed right away and only picked up a few. as far as the cross realms are treated in the client file which is uploaded they are just standard realm/faction updates.
One of the questions I had for Rollie was the TimesPlus part of the upload, I also assumed that it somehow had a data checking feature which would be mucked up by the cross realms addition.
Rollie's response was that he only used that data for the activity graphs shown here on the site. But then he did qualified his response with "at least that is what I think I remember"
Interestingly, site data conflicts with your replies...
I see people from from EU-Runetotem and EU-Thunderhorn Horde every day I play on EU-Stormrage Horde... and both of those realms list a big fat zero on faction updates for Horde on server submissions page.
And on the other end... EU-Scarshield Legion has 394 Horde updates despite the fact that there are rarely more than 60 characters online. If there was no counter of any sort... that value should be around 20 to 30 from 2 or 3 submission from me / currently available day (due to massive backlog we have).
Lastly, as this might well end as the first post of the year here... Happy new year to all!
well then that appears to indicate that Rollie remembered incorrectly and that might be a part (though I expect minor or nearly all submissions would end with E status) of the slow processing.
I haven't any data yet myself to work with.. since my uploads haven't been processed. (I kept renamed copies of the first 10 or so uploads.
And at the rate the uploads are processing.. my Dec15th upload won't get processed until mid January..sigh
Last processed file was uploaded on 2012-12-11 16:36:49.
Current server time is 2013-01-02 03:08:41
Actually... judging by that... we'll reach December 15th some time late February, early March... UGH. Maybe, as a stopgap of sorts, data expiry should be increased to 60 days? We are already close to a situation where data is "outdated" BEFORE it finishes processing.
Actually... I am still using the "old" version (pre-TOC change)... and to my knowledge, there has not been any change on my part at all... many files are variants of previous ones, just pruned. Some of them work, others don't. As could be seen, there were many successes, but as noted:
Dec 11, 12 17:50:43 0 0 E
failed, reason yet again unknown. Now the queue has gone to the area where I have no submissions, so I can't offer direct input until queue manages to reach:
Dec 30, 12 15:21:45 0 0 N
which at current speed will take a couple of weeks, at least.
I am sorry for not being able to help you and Rollie... but to be honest, such random fail / succeed pattern is beyond my understanding... as is the fact that only I and Padanfain have reported failures. We submit a lot... but I have gone above 700 000 updates in a month in previous years and the system didn't break. Now many of the files are much smaller and they still fail. *puzzled*
Dec 12, 12 17:14:13 921 2278 Y
Dec 12, 12 16:15:00 1797 3966 Y
Dec 12, 12 14:53:30 549 1753 Y
Dec 12, 12 14:07:16 469 1895 Y
Dec 12, 12 13:11:17 6231 16796 Y <--- After this submission was when I started purging after each submission.
Dec 11, 12 21:44:48 0 0 E
Dec 11, 12 19:56:45 0 0 E
Dec 11, 12 17:51:06 0 0 E
Dec 11, 12 16:39:54 0 0 E
Dec 11, 12 14:54:08 0 0 E
Dec 11, 12 01:09:55 2208 5361 Y
Dec 11, 12 00:19:07 0 0 E
Dec 10, 12 22:55:17 763 2661 Y
Was looking bad with only 38% of Decembers processed files being successful. Up to 44%.
What I'm wondering is if there might be some incorrect left over data in the file from pre CP 6.0 version
But what that could be I haven't a clue..
the only difference other then the obvious new race/class info is
the version number and the number of fields in the TimesPlus data lines
If your submissions had any data from Pre-CP 6.0 the older TimesPlus data would have 2 fewer fields in the data line. This is something that was brought to Rollie's attention for his class activities reporting.
doing a serverprune doesn't if I remember clear the TimesPlus data.. so you should look to see if you have anything from Pre MoP days in your file.. if so you should erase the file after doing your upload and have CP recreate a clean file.
If you count the fields, you should find one for each class in the game and a trailing three for the original battlegrounds.
I wish I would have kept copies of files from around the 1st of Dec but didn't
start that until the 12th. I have all submissions from then on saved away if I
have more errors in case someone needs to look at them. I never did a
server prune but around the 5th I pruned to 7 days, then to 4 days and by
the 11th I had pruned to 1 day. I am certain the 5 errors in a row on the
11th were after a /CensusPlus prune 1.
Dec 15, 12 23:56:23 0 0 N
Dec 15, 12 21:22:46 0 0 N
Dec 15, 12 18:24:47 0 0 N
Dec 15, 12 09:01:32 0 0 N
Dec 13, 12 22:33:53 0 0 N
Dec 13, 12 21:01:30 0 0 N
Dec 13, 12 18:55:40 0 0 N
Dec 11, 12 22:35:54 0 0 E
Dec 11, 12 22:16:12 0 0 E
Dec 11, 12 18:52:47 0 0 E
Dec 11, 12 18:35:17 0 0 E
The system does not like me but yeah, I hadn't done any pruning so I'm not expecting anything.
Dec 11, 12 17:50:43 0 0 E
Dec 11, 12 10:19:01 1273 3776 Y
Those two are essentially the same file, the failed one is about 0,6 MB bigger and they are both under 5 MB. And I didn't spot any entries before December 10th... and I did spot both monks and pandaren, so data as such should be ok. Why one works and the other doesn't... simply isn't apparent. Size is not decisive either, because some of the files from, say, Dec 8th and 9th were bigger.
The only semi-obvious thing is that both Padanfain and Korwynd also had failures on the 11th. But like me, Padanfain's list also has a successful submission on 11th.
I am starting to lean in the direction of the problem being in the processing end. And that, unfortunately, is entirely Rollie's area. The only thing the failed fails seem to have in common is that they come from similar time points.
Honestly, it wasn't that I was feeling left out, but now I have a failed file as well.
Dec 13, 12 02:25:19 1457 3854 Y
Dec 12, 12 21:55:11 0 0 E
Dec 12, 12 03:55:19 2587 6768 Y
That's the first fail that I've had.
On the up-side, I started keeping my files a while back so I have the file and already know which one it was. Nothing jumped up and screamed "this is the problem" when I looked at it, but if anyone else wants to take a look at it you're welcome to it.
As for old data, I used to always delete my files after uploading them, now I drag them over to a folder named with the date, but in either case the addon has to create a completely new data file next time I log in, so there is no chance of it being old data still hanging around in my case.
Dec 13, 12 00:23:25 0 0 E
Dec 12, 12 23:27:26 0 0 E
Dec 12, 12 21:32:09 0 0 E
Dec 12, 12 20:14:03 0 0 E
Dec 12, 12 18:56:45 553 2213 Y
Dec 12, 12 17:14:13 921 2278 Y
Dec 12, 12 16:15:00 1797 3966 Y
Dec 12, 12 14:53:30 549 1753 Y
Yep, I'm convinced it's a processing problem. Those were all taken after
purging. Of the errors the smallest file was about 300 KB and the largest
about 900 KB. 59% failure rate for the month so far is convincing me to take
a break.