No access to site all day yesterday?

[NHN]

Nah on a serious note, is there going to be any real changes to the sites hosting as its becoming fairly unreliable, I dont think anyone even mod/admin would disagree with that

 

[voorhees]

Nah on a serious note, is there going to be any real changes to the sites hosting as its becoming fairly unreliable

whats the times its been down this month ?
I think "fairly unreliable" is simply not true

 

[quattrojames]

It is down far more than any other forum I visit. Its been very good lately, but for the past couple of weeks, it has suffered several outages in fairness.

 

[NHN]

In comparison to other forums, I'd say its been fairly unreliable since a year ago.

Instead of saying oh its been down only so much this month, why cant there be something constructve done, so it doesnt actually go down hardly at all or have issues with pages loading, double posting etc, its either one or another, dont think that can be denied at all.

Surely Rob of all people is getting ****** about the site having problem.

 

[FactionOne]

Instead of all this cloak and dagger crap with the server specs and vauge mentions of "moving stuff around" and "optimising" why not actually put some facts up here?

What CPU and RAM is the server running?
What OS?
Is it a VM or a real box, and if its a VM, do YOU own the real box or are you just paying some colo provider for it.

An answer like this isn't normally my style, I'd rather keep to just the relevant facts; but if we're having trouble getting to the parameters of that range, answering your salvo kills two birds with one stone I suppose.

So, the reason I've not gone into extreme detail about this and that is almost overwhelmingly simple...

If "it" is broken, your knowing whether "it" is:

a) A 386 with 1MB RAM and a 40MB HDD located on the dark side of the moon and connected via two plastic cups and a bit of string

b) A Cray with 1000+ processors, enough RAM and disk to cache the genetic code of every single organism in the universe, and connectivity by means of optical fibre direct to God

c) Anything between (a) and (b)

...Makes not one iota difference to the fact that "it" is broken. Cloak and dagger it is not; containing no more detail than actually matters, it is.

To put things into context, i do some consulting for a firm in scotland who last year wanted me to arrange a dedicated box for them. For £60 a month i sorted them out with a dedicated box (real, not a VM) which has a Core2Duo E8400 CPU, 4GB of ram and a pair of 500gb disks in raid1

That box has never been down once in the time its been running. Bar reboots for updates and suchlike its been completely solid, which is exactly what you expect from a dedicated machine.

I have to wonder what exactly is going on here. I understand the site uses lots of resources, but if a 3ghz dualcore and 4gb of ram isnt enough resources then somethings seriously wrong with the forum software or the database engine... Unless ofcourse its not actually got 4gb of ram and a tasty CPU, and people have either skimped on the box to line their own pockets, or have lumped this site onto the same box as is running other sites...

The box is comfortably over specified for what we need it to do at the moment, and has more than enough theoretical headroom in resource terms to cope with the site as it is now, and following planned future expansion (and just in case you were going-to, don't panic, we're not going to be rolling anything new out until we've got to the bottom of the current issue). The fact is, that something is broken...

So, something isn't working right, the site falls over. We have to examine what evidence we can and make a decision on what the likely cause was, and work on a resolve. If it doesn't fall over again, great. If it does, we need to carry on working toward highlighting the ellusive root cause. That's where we're at.

Also, thanks for the rather cute insinuation that the folks behind the scenes are creaming loads of money out of the site while giggling at the problems EVERYONE has been having with it over the past couple of weeks. I'd almost go as far as to say I wish that was the case. None of the folks who run the place on a day to day basis have made a brass bean out of it; and not that we're bothered, but it probably owes a couple of us a few quid.

Its also ****** rediculous that only one person actually has sufficient access to fix anything when it goes wrong. Thats the number 1 rule in sysadmin. What if he gets hit by a bus tomorrow? Proper monitoring software should mean that if the site drops offline, you get an email within 5 minutes and can poke it back to life there and then. Even if it means just rebooting it and doing the investigation later because your busy, it still keeps everything running.

OK, I guess following discussions at the weekend you're perhaps better informed about the situation here but basically there are three people who can do things about issues, but unfortunately only 1 (or 1.5 at best) can have any impact on the most serious problems. That's something I'm working hard to sort out, but things will change only as fast as they do; rest assured though that delays aren't for want of trying.

I'll give you a refund of your membership fee if you want

Book-worthy!

lmfao, touche, dont worry about mine, just the sponsors, lol

Thankfully Nigel, we're in the happy position that most of our sponsors appreciate we've got our work cut out at the moment, and have faith that we're trying to get things sorted out as best we can. The sponsorship system itself is something it appears I'll need to address, following my having a closer look at the existing setup recently.

In comparison to other forums, I'd say its been fairly unreliable since a year ago.

Instead of saying oh its been down only so much this month, why cant there be something constructve done, so it doesnt actually go down hardly at all or have issues with pages loading, double posting etc, its either one or another, dont think that can be denied at all.

Surely Rob of all people is getting ****** about the site having problem.

I'm extremely "******" about the problems we're having. I'm also trying to do what I can to get it resolved satisfactorily. The fact is that I'm having to deal with technical probems at the same time as organisational reform, as well as perform many other functions; and to get that done I've got to deal with a team of others who are in the same boat.

I can't see that it makes much difference to the overall situation, but you might just sleep better tonight if you know. I'm going to be in conference with the other two key members of the team tomorrow to schedule my latest overhaul/repair plan. I'm confident it's comprehensive enough to deal with any problem areas; it's just going to be pretty labour-intensive to implement.

Again, to all who are being very patient with us, my deepest and most sincere thanks. Again, I'm just as frustrated by these issues as anyone, and trying organise a long-term resolution as speedily as possible.

Regards,

Rob.

 

[crypric23]

Just a suggestion , but when the site is down ( lets not hope it is needed ) but can you put a default page that shows something like.

" site down for maintenance " with an ETA

At least then us little people know that something is being done and do not have to just randomly check back.

I am understand there can be problems beyond everyones control , but a default page is better than simply not knowing anything.

Apart from that Great site !

 

[NHN]

None of the folks who run the place on a day to day basis have made a brass bean out of it; and not that we're bothered, but it probably owes a couple of us a few quid.

Book-worthy!

Thankfully Nigel, we're in the happy position that most of our sponsors appreciate we've got our work cut out at the moment, and have faith that we're trying to get things sorted out as best we can. The sponsorship system itself is something it appears I'll need to address, following my having a closer look at the existing setup recently.

I'm extremely "******" about the problems we're having. I'm also trying to do what I can to get it resolved satisfactorily. The fact is that I'm having to deal with technical probems at the same time as organisational reform, as well as perform many other functions; and to get that done I've got to deal with a team of others who are in the same boat.

Fair play to yer Rob :icon_thumright: well answered tbh & for the record Jase, great comeback, made me chuckle, nothing meant ref sponsor comment, just me & Jase joking, well I was

PS. If you do this for free, then I wouldnt begrudge you a few beers of the site money for the effort, not that its my business anyway, as I do with supplying genuine vcds, afterall it is our time & nobody can have a go for couple of cold ones surely, especially since you're doing them a favour.

 

[aragorn]

crypric: Its not really feasible to have a "site is down" type message unless the outage is planned. If the server dies unexpectedly, then its not really possible to put a message up, as there wouldnt be anything there to serve the message. You could stick a reverse proxy infront of the main webserver, and assuming it was running on a different host and didnt go down with the main site, you'd still end up at the same "its not working" point, albeit served from the reverse proxy rather than your web browser.

rob: The whole reason i ask the spec is to gain an understanding of whether the system IS actually comfortably specified to do the job. You tell us it is, but provide no evidence of such. It could be failing because there is some random fault or configuration error, or it could be failing because the system is overloaded (like the previous host with its memory issues).

I'm not insinuating that yourself or any other moderators are creaming anything from the site, as you arent the people behind the scenes, your sat on the front line, and i'm fully aware of the effort yourself and others put in. It seems to me though, that the people (person?) with the keys to the till so to speak have no real interest in the day to day running of the site at all, meaning its either a pain in their side (in which case they should be handing over the reigns to someone more interested/involved) or a cash cow that they're keeping ticking over for that reason alone.

I know your working hard to try and sort out the access issues, and hopefully you'll get something workable nailed down in the not too distant future. My point still stands though that you really should have a team of people to maintain the server itself. Going from one person who cant be *****, to one person who cant be ***** and one other person, just ends up meaning it dies when the new person is busy and you're in the same boat as we are now.

I realise your doing this in your spare time, but having 5 people that can fix things rather than one, means that theres less pressure on any one member of the team, and as such things get sorted quickly and more effectively.

 

[johnnythepie]

I havnt got time for your stupidity, kindly move along & go find this person norman

still ******* myself!!A master at work

 

[FactionOne]

rob: The whole reason i ask the spec is to gain an understanding of whether the system IS actually comfortably specified to do the job. You tell us it is, but provide no evidence of such.

Well I guess if you and I had enough communication before this thread that wouldn't be so much of an issue. I'll not sit here blowing my own trumpet but I can assure you I've enough time under my belt in IT hardware to tell a good computer from a bad one. In fact there's a fair chance that I was involved in the early testing/integration process of more than one component/technology in the computer you're posting-from Anyway, the server has more than enough thinking time to cope with ASN as is, and for the next three development/expansion phases at least.

It could be failing because there is some random fault or configuration error, or it could be failing because the system is overloaded (like the previous host with its memory issues).

Aside from my traditional doubts about "random" faults in logical systems, you're right, it could be any of those things. If it's the latter though, it's because of another underlying fault - as above, the box has more than enough horsepower to cope with things running normally; if there's a memory leak from something, or similar along those lines, then it's a possibility. We're debugging that at the moment.

I'm not insinuating that yourself or any other moderators are creaming anything from the site, as you arent the people behind the scenes, your sat on the front line, and i'm fully aware of the effort yourself and others put in. It seems to me though, that the people (person?) with the keys to the till so to speak have no real interest in the day to day running of the site at all, meaning its either a pain in their side (in which case they should be handing over the reigns to someone more interested/involved) or a cash cow that they're keeping ticking over for that reason alone.

I know your working hard to try and sort out the access issues, and hopefully you'll get something workable nailed down in the not too distant future. My point still stands though that you really should have a team of people to maintain the server itself. Going from one person who cant be *****, to one person who cant be ***** and one other person, just ends up meaning it dies when the new person is busy and you're in the same boat as we are now.

I realise your doing this in your spare time, but having 5 people that can fix things rather than one, means that theres less pressure on any one member of the team, and as such things get sorted quickly and more effectively.

Again, I can assure you point-blank that ASN is a "cash cow" for nobody. I'm not happy to comment in depth about issues which pre-date my administration, as that's not fair to other people. I will just say that I have some work to do with regard to renewing arrangements which provide revenue to the site which we can use for future investment (not so much for 'plodding along' with maintenance/general up-keep, but more for development past the next three planned & paid-for major developments).

As far as redundancy is concerned, we have it on three levels at the moment, but unfortunately there isn't parity at each point. We're working at getting that resolved but basic logisitcal nightmares have meant that alone hasn't been (/isn't being) easy. It's further-complicating, and damned-unlucky that we're having to fight these fires at the same time. But again, progress is being made today, and while I can't guarantee that the problem will be completely ironed-out today, I am hoping to have it isolated, and dramatically improved response times should we have recurrences. Once we're there, there's more time for us to consider taking the next step from 'safe' to 'solid'.

Regards,

Rob.

 

[Dazmo]

Ive struggled alot the past few weeks getting on this site, i type the address and the page just doesnt load, nothing comes up and i have tried from home and work and its the same story.

I should point out that over the past few weeks ive been unable to gain access to this site on at least 6 occassions.

So who knows whats going on.

 

[02audis3]

Also same with me

 

[voorhees]

try reading the thread you have posted upon...