VM Newsgroup Retention

[HIM2010]

can someone please help me understand why and what happened?

Ive been using my free virgin newsgroup account, which only allowed 7 days.
I recently signed up for a new newsgroup reader account, and downloaded some files which I hadn't noticed but were actually 500 days old.
This worked for 2 days like this, then all of a sudden, bam, I could only find articles 7 days or less.

what happened?
did i somehow get into a virgin backdoor where they have longer retention and they saw i was using it and removed me?
is it possible to get this back to longer than 7 days?

cheers!

[broadbandking]

Some articles are on the server longer, but the alot are incomplete that are older than 7 days, so I have heard.

[Kymmy]

Posts moved to new thread as off-topic within customer data retention thread

[gazfan]

As far as I know retention works as if there is a bucket for each group.

Each bucket has a limited size, but some fill up faster than others.

If you choose files from a bucket (group) with minimal traffic you may stand more chance of finding older posts.

Files from a group with heavy traffic (high numbers of large files) will be constantly running out of capacity (overflowing) & each new file posted will always displace an older one.

Files from text groups will usually have much longer retention than files from binary groups, simply because they tend to be smaller.

[RainmakerRaw]

As far as I know retention works as if there is a bucket for each group.

Each bucket has a limited size, but some fill up faster than others.

If you choose files from a bucket (group) with minimal traffic you may stand more chance of finding older posts.

Files from a group with heavy traffic (high numbers of large files) will be constantly running out of capacity (overflowing) & each new file posted will always displace an older one.

Files from text groups will usually have much longer retention than files from binary groups, simply because they tend to be smaller.

Not quite. High-end NSPs (including the Highwinds group, who VM outsource their Usenet supply to) have 99.9% completion and retention across all groups. Basically that means that so long as you're inside the advertised retention (eg 720 days) you will be able to download the articles in question. Groups do not run out of 'space'. ;) Otherwise you'd have NSPs unable to advertise x days retention, but rather a huge table describing which groups have what retention.

However, from what I heard (and this is basically hearsay) the reason for the variable retention on VM's named server is because of the fact they outsource to the Highwinds GROUP. Sometimes you connect to x reseller within the Group, other times you get y reseller. Apparently some of the resellers have not set the 7 day retention cap for VM customers, others have.

Bear in mind Highwinds resellers have over 700 days retention, the 7 day limit is a VM thing and as such is an imposed hard cap; very likely to reduce the per-subscriber fee from Highwinds to supply the service. Some resellers seem to have the cap implemented, others not.

Personally I wish VM would switch their Usenet supply to Astraweb, ReadNews (Blocknews.net / UsenetNow.net) or even Giganews and keep the full retention. Now that would be awesome. :D

[HIM2010]

thanks for the indepth info,
ok so youre saying it was luck... ok ill do a test tonight of the same files, to see if i can still get the 500 day old files... if not, then all your info does not explain how I was able to d/l 500 day old files, for 2 days, before going back to 7 day files...

[RainmakerRaw]

thanks for the indepth info,
ok so youre saying it was luck... ok ill do a test tonight of the same files, to see if i can still get the 500 day old files... if not, then all your info does not explain how I was able to d/l 500 day old files, for 2 days, before going back to 7 day files...

Why doesn't it? For two days you'd have been connected (routed) to a reseller who apparently hasn't implemented the 7 day VM cap (for example, Thundernews); and then the next day you were routed to another reseller (eg Newshosting) who do. It would explain your findings perfectly. :confused:

[HIM2010]

well as I said, Ill try to get the same files, which should be from same host right? and if it works, then what you say makes sense...

hmmm, anyway to control which hosts we use? ;)

[RainmakerRaw]

well as I said, Ill try to get the same files, which should be from same host right? and if it works, then what you say makes sense...

hmmm, anyway to control which hosts we use? ;)

No it's supposed to be random according to what I read; meaning you just take your chances and the results won't necessarily be replicable. As I said it's only what I've read before but it makes sense.