Cache problem

[cathaven]

I recently noticed, after attempting to load headers on a huge group & eventually deciding it wasn't of sufficient interest to be worth the effort, that I wasn't getting all the new headers I should. I soon discovered that the cache setting was getting stuck at around 200/200 (226) It varied a little, but something was jamming it. I disabled all the recently created anti-spam filters. I compared my group list with the spool & found several groups that had been deleted were still in the spool. I deleted them, but discovered that the huge one I'd been messing with refused to delete until I closed Newsbin, as it was in use. I closed Newsbin & deleted it, but as soon as I reopened Newsbin it was back in the spool. Evidently Newsbin hadn't finished processing it. As far as I can tell it's effectively blocking any groups that come after it in the list. It's possible that being patient & leaving Newsbin open for a very long time would result in the cache finally being cleared, but it looks like that could take days, if it works at all. Which database files do I need to delete to fix this & is there any chance it won't result in be losing posts on some groups?

[Quade]

I deleted them, but discovered that the huge one I'd been messing with refused to delete until I closed Newsbin, as it was in use.

If you exit Newsbin, delete GZ files, then restart, they won't come back unless you still have the header download in your download list (which is what this sounds like). Deleting groups doesn't stop a header download you already have running in the download list.

If your anti-spam filters are targeting the poster, they probably won't help you with the latest iteration of the spam flood. You need to use a subject filter instead. Several are discussed in the header filter topic. The spammers switched to random poster address so poster doesn't work anymore (you can still use it for other posters).

[cathaven]

Deleting the GZ files fixed it, thanks.

I'll check out the filter topic. If you've worked out how to filter those spammers who use randomly generated gobbledygook in their subject lines and poster fields, I will be impressed.

[Quade]

If you've worked out how to filter those spammers who use randomly generated gobbledygook in their subject lines and poster fields, I will be impressed.

It's relatively easy with regular expressions. For example the spam posts don't use spaces so you just need to match on a long length of characters with no spaces.

[rh]

I was wondering the same thing. Thanks for that tip!

[rh]

So what flavor of Regex does NB support? I've got an expression that works in my test tool but not in NB or regex101.com.

[\._a-z0-9-[ ]]{35,35}

[Quade]

Pretty sure you need to escape the []'s too.

[\._a-z0-9-[ ]]{35,35}

Might try:

[\._a-z0-9-\s]{35,35}

Not sure you want that second dash.

[\.\s_a-z0-9]{35,35}

You'd probably need to make sure it's a "subject reject" too.