Let me preface this by saying that when I first started my job over a year ago, I was a big fan of EBSCO. Students found their interface easy to use, they were always improving their interface and offerings, and our local Rep was really responsive to our e-mails and questions. But something has changed. EBSCO made a lot of internal tech changes this summer and I was never entirely clear about why or how it benefited the end user. But that’s fine; as long as it runs ok, I’m happy. However, I started noticing that I was getting a lot more error messages and sometimes I would click on an article link it wouldn’t go anywhere (the same page would just reload). Ok, annoying, but I still got to where I wanted to be at some point.
Last week, EBSCO A-to-Z went down for at least two days (Monday and Tuesday, so I’m not sure if it was ok on the weekend). We received no message from EBSCO during that entire period. Our Systems Librarian told me that people on the e-mail lists he subscribes to were frantic, wondering if they had broken something or if other people were also experiencing an outage. Finally, at almost 5 pm on Tuesday, EBSCO A-toZ sent an e-mail to their customers stating “Thank you for your patience. EBSCO A-to-Z(r) service is restored.” No mention of what happened. No mention of what they did to fix it. No mention of how they will prevent it from ever happening again. No even “we’re so sorry for inconveniencing you.” When A-to-Z again went down yesterday morning, I called their customer service line only to hear that yes, it’s down and no, they can’t give me an estimate on when it will be back up. I asked if they had a failover server and was told that one of their two servers has been broken and the second one is having a hard time dealing with all of the traffic and keeps crashing. So what they’re telling me is that they don’t really have a failover server because they need two servers running to handle the traffic and that’s all they’ve got (on a good day). Maybe the rep gave me incorrect information, but it certainly didn’t fill me with confidence. Yikes!
Without an explanation of what went wrong, how they fixed it and how they will prevent it from happening in the future, they are basically leaving us to assume that they are unreliable and are not doing anything to prevent outages from happening in the future. And when our contract comes up for renewal in a few months, how will all this affect our decision about whether we will renew or not? I’m sure you can guess.
When companies are having problems with their servers or with their database or any other tech-related thing that affects performance, the best thing they can do is be completely candid about it. PBWiki was having some problems recently, and, while they were down, they put up a message about exactly what was going on. And once it was fixed, they went into even greater detail about how they fixed it and were extremely apologetic. And this is a product most of us aren’t even paying for!!! And for that candor, people were much more understanding of the outage. It’s that sort of transparency that makes us forgive occasional outages, because we know they are working to fix it and prevent it from happening again, and they respect us enough to keep us informed about what’s going on. Transparency is key in business (and in libraries)!
It led me to start thinking about the fact that we (at least here) never ask questions like this. When I was choosing a host for the off-campus server I run for the library, I asked all about uptime/downtime, backups, etc. When we’re contracting with a vendor (either for a database or any other online product), do we ask questions like that? Do we ask them, “how many hours have you been down in the past three years?”, “do you have a failover server(s) and is it located geographically distant from your other server(s)?” Nope. And maybe we should. Because without A-to-Z, it is completely impossible for us to figure out which database a specific journal will be in (unless we already happen to know). So when I was trying to create persistent links for readings in the graduate program courses and A-to-Z went down, I was left totally unable to do my job. When so much of what we do is dependent on our online resources working, we should be asking these questions.
If a technology that you’re responsible for goes down, don’t just hope that no one will notice. Be candid about it. Show your humanity. Give people information. Because it’s a lack of information that will lead people to assume the worst. Saying you’re sorry is a very powerful thing. Being real is a powerful thing. It can give a face to a faceless company and make your customers much more understanding of the occasional problem.
Update: By the way, on a slightly related note, has anyone ever been able to get EBSCO RSS feeds to work? I was talking with Paul Pival today and he has been having the same problem with EBSCO, which is that we create a feed following their instructions and never get any content out of it. Mine was a keyword search for “international relations” so you’d think there would be one or two or several hundred articles on that subject. I’ve tried it numerous times in numerous different ways and have never gotten anything out of the feeds they’ve given me. Paul has also created numerous search feeds with EBSCO as well and has never gotten any content. Has anyone had any luck with it? Is there some magic trick we’re missing?
I have successfully made and used a number of feeds from EBSCO, though I have found that some aggregators don’t like them. I posted exactly how I created feeds from EBSCO here: http://davidrothman.net/2006/08/30/how-to-generate-rss-feeds-from-ebsco-medline/
However, I really dislike the needlessly complex steps it takes to generate a feed from EBSCO. That interface needs work.
Comparing PBWiki and EBSCO is a little unfair since PBWiki does not depend on paid subscriptions for their survival. EBSCO has more difficulty being transparent because their competitors would just love to find ways to outmode their product.
Does that mean EBSCO ought to use a PBWiki strategy instead? I wish. Open access would be great and I can’t wait for that egg to crack. But with the current way databases are operating, I don’t think we can expect too much transparency in ops.
I’ve also tried to use an EBSCO feed without much luck. I haven’t looked at it for awhile but I’d be interested in comparing notes.
I have had some success (knock wood) using EBSCO RSS feeds relating specifically to journals; I have *not* tried to make one work for a keyword search…
Basically, the RSS feeds seem to work fine as a Table-of-Contents service (I have been contacted when new issues for the Virginia Quarterly Review and Texas Monthly have been indexed).
PBWiki is a for-profit company just like EBSCO is. They need a competitive advantage over other wiki companies (like WetPaint, SeedWiki, etc.) too. I\’m not suggesting that EBSCO should be free or open access. I\’m suggesting that EBSCO be honest with their customers about problems they are having with their servers and to tell us how they are working to improve their systems so it won\’t happen again. I\’m not sure how that would outmode their product. It\’s much worse for business to not be transparent in areas like this. EBSCO went down again this afternoon (while I was trying to answer a reference question via IM) and it leaves all of us here with a bad taste in our mouths because they will not even tell their customers what\’s going on or apologize for bad service.
Thanks David and Adam. I have tried to use their feeds in Bloglines and syndicated on a page using Feed2JS, but so far I have not had any luck. They were keyword search feeds and I went through the same steps David had demonstrated in his post. And I ended up with nothing.
Not as big a concern to me as getting their products to work period, but I will try to create a search feed again in the future.
I’ve created both keyword and table of contents searches from EBSCO databases without any problem, and viewed them in Bloglines. Not sure exactly how I did it, but it didn’t seem terribly difficult (and I’m surely not as technically adept as Meredith is).
I was using EBCO’s Australia/New Zealand collection – possibly that database works differently from others – though I can’t see why it would…
I’ve had success with keyword feeds from EBSCO in Netnewswire. In fact, I was getting so many results unsubscribed from it.
Thank you for venting — you are SO right… I’ve been complaining about the way they’ve done some of the upgrades this summer lots of places… not to mention the way they mounted the Inspec Archive (ARGH!). I’ve been getting these errors, too, but not with A to Z. I was just having such a crazy week I didn’t stop to figure out what was going on. I also assume that no one will tell me… they only tell the office where the check comes from which is an hour away. Now I’ll go back and grab that post I’ve had in draft and get it posted!
I haven’t played around with the EBSCO feeds yet, but have noticed a definite upswing in database problems in the past few months. In fact there was one yesterday. Luckily we get all of our EBSCO dbs through a state-wide consortium so we don’t have to deal with them directly with EBSCO customer supprt, which doesn’t sound too fun!
thank you. thank you. thank you and bless your soul. We have felt crazy and cursed in our library, and have also been made to feel that it must be on our end (ie filter, firewall, or bandwidth). We updated all the ips and url changes on our side (and seriously – they send me enough emails, they could have sent me that email too instead of it being after the fact). Our tech services guy Brad has been very nice and patient and somewhat persistent – and actually admitted he just didn’t know in the end. The problem is just so durn intermittent its hard to pin down what is happening.
arg – i meant to quote a section of your post above, to explain what I was referring to:
“sometimes I would click on an article link it wouldn’t go anywhere (the same page would just reload).”
lol! I’m also glad it’s not just me who is noticing this problem. 🙂 I’ve also seen a few people complaining about the EBSCO page loading weirdness on the EZ Proxy listserv. Of course, no one has gotten a satisfactory answer as far as I know. Sigh…
I don’t have an answer to your question, but if you’re ever stuck not knowing what database a periodical is indexed in, try Ullrich’s (in paper or online). That’s one of its main purposes.
I’m going to come across like a Gale groupie here, but when our Gale databases go down, even for fifteen minutes, we get immediate e-mails from tech support telling us what’s going on (including the cause and estimated re-up time), e-mails letting us know about the progress if it’s an extended outage, and e-mails letting us know when access is restored. This is something I greatly appreciate as the tech manager for our library, and counts as a checkmark in Gale’s favor for future database purchasing decisions.
Things like these are why you (or at least the provider of the service!) should always use some kind of monitoring service for anything you consider important. That way they will know right away when something goes down the drain.
I had a site stop working for me last summer and didn’t know about it until almost a day later. A total disaster… After that I started using a site monitoring service called Pingdom. Has saved my bacon on two occasions since then. (I know, I really should switch to a more reliable host as well…)
Hey Meredith, after having a conversation with David Free I think I figured out our problem! It seems the EBSCO feed doesn’t do anything until a day passes and the search is run within the database, which is counter to every other RSS feed I’ve plugged in to my aggregator. I guess I expect to see something right away to verify that it’s working!!! Anyhoo, go ahead and leave an EBSCO feed in your aggregator for a couple of days and see if it doesn’t mysteriously spring to life…
I’ve been struggling to implement our A-to-Z list since last summer. One of the major questions I have is related to our backup. EBSCO customer support told me that by downloading our collection on a regular basis, I am creating our backup copies. However, I am not completely convinced. The proxy settings in the download list do not match the settings in the Collection Editor, and EBSCO could not answer my simple question “why is that?” I am also not convinced by them that all my notes are somehow miraculously saved in the download excel sheet, since I do not see any columns where they appear. I wonder if anybody had used those downloads as the backups, and if they really store all the information you have in your Administrator module. If you know of any listserv(s) where I can post that question, please let me know.