Write Records tossed an Exception

[skids]

v6.40build 2059 Astraweb

I am almost constantly getting a "ERROR NNTPServerWorker - Write Records tossed an Exception". I restart Newsbin and it works for a while and then the error starts again. The message shows up 12 times per minute. After the error starts no headers are downloaded and nothing else can be downloaded. For example I restarted Newsbin at 5:49 this morning and the errors started again at 11:11am and that error is the only thing in the log after that time, like I said, 12 times per minute over 5 1/2 hours. The problem started last Friday. I did no updates and added no groups and made no changes to settings of any kind. I have 3 servers set up and only download headers from one of them. I have changed my download server but it didn't change anything. My original server was using SSl but the one I'm using now is not.

I've searched for this error and found 3 references, one in the v5 forum from Quade said it was a major problem and to delete the group and re-download it. How do you know what group? There is no reference to a group in the error. After a restart a couple of groups keep downloading some of the same files over and over. Could these be the problem groups? I know this next part is your problem but I live in the sticks and have wireless service. Deleting all the groups and redownloading is not really something I want to do.

[Quade]

This isn't likely to have anything to do with file downloads.

In the data folder, in the spool_v6 folder, if you search, you'll find RV4 files. It sounds like one or more have become corrupt preventing the importing of records into the group. If you can figure out what group it is, you could probably delete the RV4's from that group. That would be best case because the RV4's will need to be re-downloaded and that means re-downloading headers for the group. If you can't figure out what group it is, you could delete them all. Hopefully you won't lose much data.

[skids]

Ok, I'll do some digging.

Question? I was only downloading headers from one server. By switching servers and making the new one the download server, that forced me to download all headers again. Would that have created new RV4 files or used the old (corrupted) ones?

[Quade]

Typically headers from one server work fine on another so, switching servers doesn't require you to re-download all the headers again. I'm going to guess a file is simply corrupt and there's probably nothing you did that actually caused it.

I'm actually getting rid of RV4 files in 6.5 in order to speed the feed of headers to the databases.

[skids]

I had no choice when I switched servers. Newbin automatically downloaded all headers for all my groups all over again. I didn't do it, not on a wireless connection.

On my original problem, I disabled the groups in one of my container folders and the error seems to have gone away. Now to figure out which group in the folder is the culprit.

[Quade]

I had no choice when I switched servers. Newbin automatically downloaded all headers for all my groups all over again. I didn't do it, not on a wireless connection.

It's a function of the "Download Age". This determines how many headers to download the first time a server is visited or a group is added.