Ik ben het eens met Funcracker, veel mensen reageren nu ineens wel erg overdreven. Tuurlijk is het vragen van een donatie vaak geen goed teken, maar hoeft niets te zeggen. Bij Seti vragen ze ook om donaties. Er wordt niemand verplicht ook echt iets te doneren, sterker nog Jay rekent er niet eens op. Maar alles wat wel komt is meegenomen. Ik heb gisteren een gesprek met Jay over een aantal zaken gehad en hieruit bleek, even kort, dat ECC2 als project zeker niet in gevaar is. Ook het aanbieden van een snelle server achter een snelle lijn helpt niet.
Ook nog even als reactie op Erik Jan. De
stats worden gehost door Innuedo op een 100Mbit UTwente lijn. Maar statsserver is iets anders als de DP server, die dus werkelijk de dp's ontvangt en verwerkt. Die staat bij Jay thuis op een gehuurde sDSL lijn (1 Mbit).
Zal nu even wat stukjes quoten uit ons gesprek:
[21:01] <ColdFusion> i saw that ecc2 started to ask for donations on the website
[21:01] <JBerg> Yes, about a week ago.
[21:01] <JBerg> Isn't a critical situation. But things are tight for everyone - including the ECC2 project.
[21:02] <ColdFusion> is that for better hosting? extra bandwidth?
[21:02] <JBerg> Mostly the bandwidth situation.
[21:03] <JBerg> I'm paying a premium price ($260/month) for a high quality commercial line. But since I'm not employed, it's been difficult to continue paying for it. I'm seriously considering converting to cable. But if I do, it would be a major impact.
[21:04] <JBerg> Reliability would suffer dramatically. Upload speeds would choke.
[21:04] <JBerg> But how do I continue justifying paying so much, when the only need for the high quality is the ECC2 project? If it wasn't for that, I could switch to cable tomorrow and not have a problem.
[21:05] <JBerg> In addition, I've been needing a UPS server for the DP server. It's the only box I have that isn't protected. But I don't have any spare UPS boxes for the server.
[21:06] <ColdFusion> will the future of this project be in danger because of this? say if no donations come?
[21:06] <JBerg> The project will continue. But, I may end up moving eCompute to a cable connection. If I do, expect outages. Also expect uploads to start choking. And the email submission and manual uploads may stop.
[21:07] <JBerg> The email submissions and manual uploads are done through the eCompute mail service. If I go to a cable connection, I can't run a mail server.
[21:07] <ColdFusion> and say: i got a server and good reliable line for you for free, would that help?
[21:08] <JBerg> There's no way we could operate the project with a remote DP server. And frankly, moving the DP server would require it to be off the air for a week (or more).
[21:08] <JBerg> It's far more than just a web or ftp server.
[21:08] <JBerg> It requires more than a box and a connection.
[21:09] <ColdFusion> i see, so ppl offering you a server and a free line, wont exactly help?
[21:09] <JBerg> Not really
[21:10] <JBerg> Remember that the database is a custom one designed and written by me. Also that it uses some very special tools that I've developed. And it requires about 2 hours a day normally to maintain everything, as well as a 16 hour backup cycle once a week.
[21:13] <JBerg> I understand where the offers come from. Everyone wants to help. But frankly, the DP server was never designed to be "portable" or to allow remote operation.
[21:32] <JBerg> The project will continue. But if nothing changes financially, there is a strong chance that the QoS of the project will degrade.
[21:32] <JBerg> The QoS of ECC2 has been remarkably high.
[21:32] <JBerg> This QoS is due to certain factors.
[21:33] <JBerg> 1. I have been available 16-18 hours a day, 7 days a week. To spot/fix problems when/if they arise.
[21:33] <JBerg> 2. I have provided a high quality high QoS Internet service "for free".
[21:33] <JBerg> 3. I have 30 years of Engineering experience to draw on, combined with 15-20 years of system design.
[21:34] <JBerg> But the fact of the matter is that I'm running low on resources. As a result, I'm going to have to find outside employment. This means I'll be available far less than I have been. I don't think this will create a major problem, but it is going to happen.
[21:35] <JBerg> Second, I am probably going to have to reduce costs. This will impact the QoS of the project. Plus it means certain resources will no longer be available.
[21:36] <JBerg> I honestly didn't expect anyone to provide any donations - and so far I haven't been surprised.
[21:37] <JBerg> But I'm looking at having to keep this running for another 6-18 months. And honestly, I can't do it at the current level of QoS. I can not continue to give it 16-18 hours a day, 7 days a week. And I doubt that I can continue to pay $260/month for this top quality connection.
[21:39] <JBerg> So far today (about 6 hours), some team has been doing a mega-flush. This has resulted in approx 4-10 DP/sec non-stop for a half dozen hours.
[21:39] <JBerg> You know what that means? That translates to approximately 1024 - 4096 bps.
[21:40] <JBerg> But in reality with the TCP handshaking, that's actually close to 25% of my bandwidth being used.
[21:40] <JBerg> With peaks of 50-75% of the bandwidth for minutes on end.
[21:40] <JBerg> Now, imagine me moving to a residential service. Think what the results would be when just one turkey does a big flush!
[21:41] <JBerg> Right now, nobody even notices. My bandwidth is capable of handling my normal operations, as well as the big flushing.
[21:41] <JBerg> You see why I bitch so much about turkeys doing big flushes when they don't need to?
[21:42] <JBerg> The only income I have, is through what I earn online. Take away all of my bandwidth for a day, and I lose a day's worth of revenue!
Ik hoop dat dit verhaal wat verhelderend werkt bij de meesten. Dat deed het voor mij over bepaalde zaken in ieder geval wel.