slow online cattery ?

[ruby Tierbal]

Tad i do love all ur time u spend explaining things and putting so much in ,, i am sooooooooo not technical on any level lol.
i try to keep cattery cats in there so i can control the times / love / days etc so im certain there has been a slip, but im watching also to be sure , with all us breeders sending tickets on a whim its must get crazy some days , we know if there is a issue it will be sorted by the team , just making people aware to see if its a bigger issue or just something little Please let us know Dev if u see anything too
thanks for all the replies once again

[Devilness Chant]

I'll let you know when they all pop, but so far everything seems normal.

[Devilness Chant]

The ones I'm testing are all at 91 percent love, and the others seem to be moving along nicely.

If I notice anything again, and you might want to try this too. Write the stats down at the same time every day. That way if we need to write a ticket, we can give exact times and info.

[ruby Tierbal]

thanks for info , yes i've done a pair too and just to be awkward as cats r at times they arrived a little early , sure they make me think i'm crazy il monitor it well away as will online have online cattery access for a few weeks thanks again!!

[Devilness Chant]

(07-13-2013 12:39 AM)ruby Tierbal Wrote:  thanks for info , yes i've done a pair too and just to be awkward as cats r at times they arrived a little early , sure they make me think i'm crazy il monitor it well away as will online have online cattery access for a few weeks thanks again!!

rofl Yes kitties and rl cats have so much in common. They all want us to think we're crazy.

I've had kitties that were like 3 days old and in my mind it's like, "that kitty should be at least a week old!" then check the parents and realize, it's the right age. I think it's cause we want those kitties to hurry up and breed! lol

[dakillakm Resident]

I have noticed lately that two cats paired for multiple times, both in cattery at all times, will end up with their love off a percentage or two from one another. Next time I notice this I will file a ticket. Not sure how or why this would happen and I can't remember exactly when I last noticed it.

[Devilness Chant]

(07-14-2013 01:36 PM)dakillakm Resident Wrote:  I have noticed lately that two cats paired for multiple times, both in cattery at all times, will end up with their love off a percentage or two from one another. Next time I notice this I will file a ticket. Not sure how or why this would happen and I can't remember exactly when I last noticed it.

I think it has to do with their sleep schedules. Unless they fall asleep at the exact same time I think that will throw them off a hair. I could be wrong but it seems like that might be it?

Or like maybe kitty A was opened at 12:00, kitty B was opened at 5:00. Maybe just enough of a difference to slowly throw their schedules off?

[Tad Carlucci]

Off the top of my head, I'd suggest it's likely that timing is everything. Not just yours, but KC's, as well.

As a systems analyst I immediately think of two ways to implement a time-based function on my web site.

The simplest, and what is most likely the go-to method for most programmers, is to use what we call "a cron job". That is, we schedule something to run, in the background, on a regular basis. Say I want to implement the "Love clock". I decide that once-an-hour is good enough. I write a program to "tick" the "clock" for every cat in the Cattery. Then I tell "cron" to run that program once an hour. (Oh, where did that come from? Cron is a time-based job scheduler in Unix-like computer operating systems. The name cron comes from the Greek word for time: "chronos".)

Now, that program can't really update everything at once. Instead, it updates each cat, in turn. Say your first cat comes early in the list. And, your other cat comes quite a bit later. And, say, it takes 5 minutes to get from one to the other. In a simple implementation, I'd "tick" the first cat and record when that tick occurred. Then, 5 minutes later, I'd tick the second. See the problem? YOU may have birthed them just a second or two apart. But my cron-based program moved them 5 minutes apart. That's this hour. Next hour system load is a bit lower (nobody's hitting the web site!) and it only takes me 4 minutes to get to the second cat. But, my ticker looks and says, "Nope, not been an hour, yet (it's only been 59 minutes) .. you gots to wait." Oh oh! The first cat just ticked; but the second won't get the matching tick for ANOTHER HOUR! It's late at night, and loads will stay low for a long time, so the cats will be 1-tick apart. Now, if I'm smart, and if loads go up and it takes 6 minutes, that second cat will finally get two ticks at once; when it looks like 121 minutes of ticks needed, instead of only 119. But, then, that might never happen because some more cats could come along and the distance between your two cats gets longer and I'll NEVER see that I missed a tick and double-tick. Of course, if I don't check, I'll never see I missed a tick. And it's SO easy to just write a "If it's been a hour or more since the last tick, tick ONCE."

Do I have to do it that way? No. But it's easy.

If I'm willing to record a bunch of extra data with each cat, I could eliminate the need for the cron-based operation altogether. Let's face it, if you never look at the cat, you don't really CARE. So what if it "should" have produced a box last week? You never looked at the cats, or the boxes, so you don't KNOW. But you're pretty sure it "should" have happened. How do you know? You hit the web site. Well, I can do all the ticking and tocking and box production, the moment you actually look! (Hey .. yep .. it's Schrödinger's cat!)

That's a LOT of work for the programmer. But it's a net win for the server since it never does any work if you never look! And, done properly, it's always dead-accurate.

So, sure, it's entirely likely that, no matter the intention, if a cron-like process is behind it, there's probably some drifting going on, at least temporarily.

[GhostCollective Resident]

*falls over due to information overload*

[Tad Carlucci]

Earlier in this thread I said I'd monitor the (totally unrelated) "You're out of food!" emails.

That's run by a cron-like object inside Second Life, so that it can send you an Instant Message (which, for me, forwards to my email).

I've been watching for a week. I know KC said it's dependent upon a lot of factors like simulator load and restarts.

The results clearly show what I'm talking about with cron-like processes. I have no idea exactly how it's implemented. But there are two messages "per day". The "PM" message shows up exactly 12 hours, plus or minus a few seconds, after the "AM" message. But the "AM" message is always 12 hours PLUS 10 MINUTES after the "PM" message.

So, if nothing intervenes to change the schedule, over the course of 144 days, when there should have been 288 messages, there will only have actually been 286.

Now, yes, this is totally unrelated to the cat functions. But, if the same problem were to appear for the "Love clock" (and, if it's cron-like, I suggest it probably does), that would definitely cause noticeable differences.