Saturday 10/5/14 dreadful day for stability ?

[OliasOfSunhillow]

Unlike previous days this day was returning garbage on price requests at numerous times during the day. The new API seems far less stable than the old API. Did anyone else have loads of exceptions occurring ?

[bnl]

yes - lots of them
/Björn

[Oscar Jack]

Me too.

I think I am going to stick with the old API until there is no choice. I think it speaks volumes that they did not "Go Live" with API-NG on the 30th April as planned and now have no fixed live date. If they have not the confidence in it why should we take the risk.

[OliasOfSunhillow]

Good to hear I am not alone. I have emailed them to comment. I am still using the old but my API-NG test app which has been running smoothly was crashing regularly. I have now introduced lots of exception handling to increase graceful degredation but that does no alter the fact that if odds are not being returned you cannot place a bet.

[BetfairDeveloperProgram]

Hi All,

We have just posted an update regarding Saturday's issue via http://forum.bdp.betfair.com/showthread.php?t=2353

Thanks

Neil

[Oscar Jack]

So a network failure primarily affecting API-NG takes several hours to fix.
listClearedOrders stops returning more results every Saturday and Sunday.
Questions on the forum directed to the Betfair Developer Program don't get answered.

Seems to me this indicates how low a priority API-NG is and therefore how much resources are allocated to it.

[BetfairDeveloperProgram]

Hi Oscar Jack,

Thanks for your feedback. You should be aware that Saturday's issue affected API-NG, the Betfair website (Sportsbook & Exchange), as well as all gaming products.

Questions on the forum directed to the Betfair Developer Program don't get answered.

We'll answer questions via the BDP forum when possible, but support queries into BDP from API subscribers are prioritized.

Neil

[Oscar Jack]

It must have been a close decision by the powers that be to not declare the problems major enough to class it as an Outage and therefore disable Betfair totally, ensuring a level playing field, since it appears the old API was ok yet trying to place a bet via the website was a nightmare.

[Dodgee]

Sorry to be blunt, but if a "hardware failure" can cause such a major issue for 12 hours on the busiest betting day of the week, that raises big questions about the design of your platform.

For years now we've had to deal with regular outages and if anything, as many others have said, the problem only seems to have got worse on API-NG. Surely some lessons should have been learned, and resilience and availability should have been put top of the list in the planning of this latest platform?

I realise I'm speaking out of turn in some ways here because I don't work for Betfair and I have no idea what challenges you had to deal with when designing this platform, but I can't help but voice my utter frustration at the apparent lack of accountability when these problems happen.

All day Saturday, all I could get out of the helpdesk was "It's being looked into, we will provide an update when we get one". Of course the update never came. At one point, I voiced my annoyance and the response was "Technical issues can't be helped". Speaking as someone who has worked in network infrastructure for the last decade, YES THEY CAN. I can think of hardly any hardware issues that couldn't be solved with some simple redundancy.

No lost funds are ever refunded in these situations, despite Betfair being at fault, and in the aftermath, the issues are simply swept under the rug until the next time. I've filled in many surveys with comments and suggestions on this subject and invited a response from Betfair but no-one has ever seen fit to contact me. The "Customer Commitment" also seemed to disappear pretty rapidly after being established, presumably because it became obvious that it was impossible to honour.

I just think paying customers deserve a lot more than contempt and apathy.

[Zhong]

Sorry to be blunt, but if a "hardware failure" can cause such a major issue for 12 hours on the busiest betting day of the week, that raises big questions about the design of your platform.

For years now we've had to deal with regular outages and if anything, as many others have said, the problem only seems to have got worse on API-NG. Surely some lessons should have been learned, and resilience and availability should have been put top of the list in the planning of this latest platform?

I realise I'm speaking out of turn in some ways here because I don't work for Betfair and I have no idea what challenges you had to deal with when designing this platform, but I can't help but voice my utter frustration at the apparent lack of accountability when these problems happen.

All day Saturday, all I could get out of the helpdesk was "It's being looked into, we will provide an update when we get one". Of course the update never came. At one point, I voiced my annoyance and the response was "Technical issues can't be helped". Speaking as someone who has worked in network infrastructure for the last decade, YES THEY CAN. I can think of hardly any hardware issues that couldn't be solved with some simple redundancy.

No lost funds are ever refunded in these situations, despite Betfair being at fault, and in the aftermath, the issues are simply swept under the rug until the next time. I've filled in many surveys with comments and suggestions on this subject and invited a response from Betfair but no-one has ever seen fit to contact me. The "Customer Commitment" also seemed to disappear pretty rapidly after being established, presumably because it became obvious that it was impossible to honour.

I just think paying customers deserve a lot more than contempt and apathy.

I don't think the customer service staffs of Betfair are well trained. They always want to skip the problem and talk as if they are working for some bureaucratic organizations. This causes big appointments when shit happens.

For example on Saturday when server is down, when I explained to the customer service that the bets were not shown on the webpage, and it was possible that user bet again and again and thus would risk more money than he was willing to if he thinks that all his bets were not settled, the customer service, without really understanding my statement, simply answered "I assure you that all bets are safe". Well, the bets are not safe!

When I asked if there's a timeline to fix the problem, he also just answered "we are looking into it" and didn't point me to any schedule or anyplace where i could check if the problem is fixed or not. And the twitter message that mentioned the server was down was 1 hour after the problem really happened.

All in all, I feel that the customer service is very cold and bureaucratic when facing the real problem. They just want to hide away from the real problem. I think people are disappointed not only due to the server problem but also due to the way betfair deals with the problem.

[Ipposnif]

I hope BF will give to the API-NG team the needed resources to bring rapidly this project in harbour.
Because i find myself in a "limbo": i can't use the old api for commercial reasons and i can't use the new api because they are not finished.