• Welcome to Simple Machines Community Forum. Please login or sign up.
October 18, 2021, 08:15:04 AM

News:

Want to get involved in developing SMF, then why not lend a hand on our github!


SMF Database Error! emails started when 2.0.4 upgraded from 2.0.3

Started by Vince S, February 08, 2013, 05:42:06 PM

Previous topic - Next topic

Arantor

There's probably a reason why he can't provide a definitive answer: because I'm not sure there is a better one than what's already on the table.

The discussion in that thread, all ultimately either revolve around 1) not using ext/mysql and checking the server socket connection (which isn't an option due to the way most hosts are configured, *especially* shared hosting), or 2) checking the connection and trying to reconnect in the event of a failure.

If the server dropped the connection, it dropped it for a reason, and that's almost always because it's overloaded, so creating *more* load by trying to throw more connections at it seems to me to be a very bad thing. But hey, if that's what SG is thinking we should do, I'm more than happy to provide a patch - but I won't include it in base SMF because I'm really not convinced that's the smart thing to do.
No good deed goes unpunished
All helpful urges should be circumvented

Kindred

See....   the thing that really suggests (almost proves) that this is a HOST SERVER issue is that in thousands of installs of SMF 2.0.4 and 2.0.5, this problem exists one one or maybe two hosts.
Please do not PM, IM or Email me with support questions.  You will get better and faster responses in the support boards.  Thank you.

"Loki is not evil, although he is certainly not a force for good. Loki is... complicated."

Vince S

Quote from: Arantor on October 04, 2013, 09:11:25 PM
There's probably a reason why he can't provide a definitive answer: because I'm not sure there is a better one than what's already on the table.

No, this is pure belligerence on your part - did you even read what he said? He says its because he can see an error with your logic but doesn't have the depth of knowledge to provide a better answer - which is a LOOOONG WAY from your position that there isn't one!!!!

Quote from: Arantor on October 04, 2013, 09:11:25 PM
If the server dropped the connection, it dropped it for a reason, and that's almost always because it's overloaded, so creating *more* load by trying to throw more connections at it seems to me to be a very bad thing.

Frankly that's totally a guess on your part. When the SG guys had a proper look at exactly what was going on for specific times and errors they found nothing that would indicate there was a core problem there. And they were not in the "sweep it under the carpet" mode you like to conveniently suggest is where they are incentivised to be.

There is a problem here. But only skilled people can solve it. So how about you do some Googling and try and come up with something better that works? Are you up to the task of actually solving the problem or is it just going to be stubborn push-backs for as long as I keep niggling away at the core issues here? So you're pretty good and know everything? Time to prove it! Or are you like the rest of us and know when it is time to recognise standing on your digs is not such a good look and ego needs to go to be replaced with a growth experience and leap forward into an even better place. There is bound to be better error trapping out there that doesn't generate all these bs errors that are supposedly dropped connections when probably only a small qty are. But are you clever enough to find it? I'm not.
Try figuring out where all this is going to keep coming from: Millionaire Baby? Or just pass me a beer and we'll sort it, thank you.....

Vince S

Quote from: Kindred on October 04, 2013, 09:28:27 PM
See....   the thing that really suggests (almost proves) that this is a HOST SERVER issue is that in thousands of installs of SMF 2.0.4 and 2.0.5, this problem exists one one or maybe two hosts.

Hardly definitive proof when most people would just tick the mgt panel box to get rid of the pesky email notifications, some maybe after checking with their hosts and being told there is no drama. Also I note earlier in this thread several others have chimed in to say they also were having these issues. Unlike most I am the only one that didn't fade away.

For me, unless my challenge to Arantor is taken up I am left as "the meat" here. I only have the choice to disable the error notifications (163 since my post a week ago) and frankly I am about to do that and would suggest anyone else receiving these emails en masse do the same. There is probably not going to be a true solution found as whatever exactly is going on it is currently beyond the means of good people acting competently to figure it out in a way that any other action would get a better result. Maybe the jury is still out on whether the error itself represents a real problem for real people as, despite the explanation given earlier that someone somewhere will be looking at a blank screen / similar, not one single report of an issue of consequence has come my way during these many occurrences. And they do let me know when other stuff goes wrong. So at worst there will be a screen refresh done and it works from there for the supposed disaffected user, which hopefully is a bot that is really triggering the issue so no-one cares/knows. Maybe this is conjecture but it is based on the facts.
Try figuring out where all this is going to keep coming from: Millionaire Baby? Or just pass me a beer and we'll sort it, thank you.....

Arantor

I'd take more credibility from what he's saying if he were actually considering the issue, except he appears to have missed the point.

QuoteWhat the is_resource function does is to check if an already established connection is active.

Yes, exactly. That's the problem. The connection has failed AFTER it originally started. That was what we changed: checking for a connection before actually trying to use a query.

This is the problem we had to deal with: not the connection failing in the first instance, but the connection failing at some point after it had started. And that's what this does.

If mysql_connect returns false, we already separately deal with that. This is not about dealing with the initial connection failing but the connection failing afterwards.

There are only two solutions to that: die gracefully or reconnect.

If it's dying after connection started, I can't see any reason for that other than a bug in PHP, a bug in MySQL or a bug in the connectors - or the server's overloaded and pruning connections. None of which we can do anything about.

Still, I was willing to be mindful of trying to help but not when you basically call me incompetent. Have a nice day.
No good deed goes unpunished
All helpful urges should be circumvented

Vince S

Quote from: Arantor on October 04, 2013, 09:44:31 PMStill, I was willing to be mindful of trying to help but not when you basically call me incompetent.

My apologies, I am not for a moment calling you incompetent, never my intention. I am simply asking that you with your skill and breadth of vision open your eyes to the possibility there is a different way, and challenging you to find it. It is a totally-other-than-competence issue what you do with that. It is more about strength of character and you just sored "a win" there by not reacting poorly to my baiting. Well done! Can you do more......?
Try figuring out where all this is going to keep coming from: Millionaire Baby? Or just pass me a beer and we'll sort it, thank you.....

Arantor

Honestly? I already did more by not quitting the SMF dev team on the spot.

As has been noted elsewhere, I'm actually suffering from depression and am frequently amazed by how much confidence everyone else seems to have in my work when all I can see is what I've done wrong.

True story: only today I had to go and look up the answer to a problem. This doesn't sound strange until you realise it was my own problem from 2 years ago that I solved 2 years ago where I'm implementing something for SMF I already did before and I couldn't even remember how to do something as trivial as what I actually did today. It was a simple matter but I couldn't even remember how to do that.

All you've managed to do here is tell me I'm not good enough to find a solution, especially as the first time you noticed there was a problem was in my code - and both you and SiteGround are telling me I'm not good enough to do better. Even if it wasn't what you meant, it's what it was. Thanks, I needed that.

I have tried, numerous times, over the last 2 1/2 years to find a better solution. The only alternatives have already been outlined, and I have stated why I believe these to be worse. Let someone else have a go because I've done all I can, and it hurts that I can't actually solve this issue once and for all.

Meanwhile, I'll try not to screw up SMF with more of my incompetence. Have a good day.
No good deed goes unpunished
All helpful urges should be circumvented

Vince S

Thank you for explaining that. I understand what you are saying and absolutely not a bone in my body wants to add to your pain.

It is a trap with the written word that different things are conveyed than what are there. A chill pill is always the best medicine and we don't even need to know the reasons why!

In my life experience I personally have heard a thousand times why something can't be done, and then I either do it or - preferably - help them to find the way. But this is in core eng'g stuff / related hobby areas. I have had Vince .vs. world experts on more times than I would like to count, and never been wrong - purely because one doesn't take such a position unless very sure of the solid ground. It is the nature of the human psyche and never ceases to amaze me how far out people will go in thin air just because they can point at a couple of things that support their view, regardless of how much substance there is to alternate views. Somehow it becomes about the contest rather than the thing - have a debate about the existence of God where the only possible conclusion is that we don't know (and even have no basis on which we might know) - but check out how many other places good people arrive at on this topic within our various societies.....!

As I often enough say, if the bs is coming from me, as soon as I spot it, it's gone! So I get to be wrong (or more typically it is miss something then re-evaluate) and am comfortable with it. This is not "failure", just "best endeavors that didn't get there". I KNOW I will think particular things that don't later pass close scrutiny, and at that point I will reassess and change direction, or look at actions to validate the current or some other view. Which is all I've been trying to do here - and my "push back" stuff comes from an observation that you have unquestioningly accepted some base assumption that I SUSPECT is invalid - but only you can determine this for sure. All I can do is push for a red hot go, and this is a loooong way from saying you are actually wrong or even that in some way refusal is incompetence. I'm just going with the evidence which I currently see is being too easily dismissed. Equally I recognise the evidence has questions about it. Which typically tells me "there is something else going on that is not currently on the table". And I suspect is going on here.

In this instance I so absolutely mean no disrespect or put downs towards you, regardless of how you may interpret what is said since your brain is maybe not functioning how it should be. Draw strength from that knowledge, it is "just you being you" and you, and the rest of us, can jolly well be ok with that! The bit where others think you are a smart cookie will have a lot of truth to it. Believe it!

Let me be frank, I had a problem with anxiety for many years of my life - and yet you would not meet a more competent or confident person (OK, maybe you would - but you know what I'm saying). I am not driven by ego one iota, but I think its cool how I often pull off extraordinary things. You seem very similar in that regard. NOW, I know what anxiety is all about and still get it - but have learned to recognise it for what it is, I know how the brain function feels and what comes next and how to adapt. I saw a shrink (when I was 43) who put her hands on the problem - as she said "your brain works so exceptionally that it has already found 4 solutions to the problem before you even knew you had one" - this is related to anxiety coping situations. Now I know when I am likely to "walk into" something that may cause anxiety so have the necessary strategies ready to go. Nothing "special" - simple awareness is in and of itself the best tool. But I find it fascinating to watch people that do exactly the same things as I may do, but I where I have to put a level of conscious effort in they appear not to. It is all part of getting comfortable with yourself, who and what you are, as there are no magic fixes / drugs that will deal with other-than-serious brain chemical deficiencies. Someone close to me was also affected with much higher range issues than I, and my personal experience was extremely helpful to them to find a way out of the mire of depression and related aspects.

Again the above is just the written word, a remarkably deficient tool for complex topics! But you are young and have your life in front of you, maybe there is a clue in this stuff that helps....

That you have put a lot of your life effort into SMF is no small thing, and you make the point so you see it. Is it giving back in ways that work for you? It needs to - but if it isn't then you need to make different choices and the rest of us can jolly well be OK with that as - ultimately - we all know why we are here and most will share that altruistic piece in our core that wants to make the world a better place but not at the expense of our fellow man (/woman).

Before we got this level of personal understanding, on the current topic you are sitting in "looks fine to me" and I am in "nice try, but can you please look at it from a different perspective?". It doesn't actually matter where it lands, but how it gets there is more important. I am perfectly OK with any version of "it is what it is, and stuffed if I can think of a better way". Storming off in a huff - even a polite one - is not helpful to anyone. Although I will admit I have now turned error reports off! Purely because I see that, whatever it is, finding a better way at the coding level is not going to come easy or any time soon, if at all. This is now a "feature" of using SMF on shared host servers (regardless of exactly what the core issue is) and the adaptive solution is pretty easy to implement. And that's OK!

And to you, have a good day - after all, it is the first day of the rest of your life!
Try figuring out where all this is going to keep coming from: Millionaire Baby? Or just pass me a beer and we'll sort it, thank you.....

Advertisement: