Beck38
07-13-2013, 11:32 PM
I know everyone tends to point to GN as the 'tops', but occasionally they 'run off the rails' too (but usually 'recover' missing data within 12 hours or less).
Last night around 2200hrs (10pm Pacific time, as they are HQ'ed in LA/California) they stopped adding to various newsgroups and continued that for around 8+ hours. Various peers that are bound 'tight' to GN (like Blocknews) fairly quickly recovered to other peers, but some (like Binsearch as a good example) never did, loosing those 8+ hours into the 'bit bucket'.
As usual, I'll bet 99.999+% of the GN (or others) never noticed, but I did (as I use GN to scan groups for 'things', and it being Friday night in particular, 'Real Time' is a 'must get' and I noticed the problem right off). Like I said, Blocknews recovered pretty quickly (in times past, they always seem to do so whenever any of the other majors like GN or Astra run a bit of the rails, shows really good programming and operation there).
Any rate, around 06:30 (630am) this morning, it appeared that GN picked things back up. In the past when things have gone 'south' with GN, a few hours later they've 'backfilled' the missing headers in those groups in a few hours, but as of now (~1630, again Pacific time) they haven't.
I'll continue to monitor through this evening, but it just goes to show that even the 'high priced spread' does occasionally have problems. FYI, the usual 'suspect' in 'bad ops' (Astra) had zero problems all throughout this 'fade' at GN, and that 'best of the fills' Blocknews, really showed that it saw the problem and picked things up virtually right away.
UPDATE: As of 2200hrs (10pm) tonight, about 24hrs past the time when the 'event' started, still no 'backfills' to GN-US OR GN-EU. I'll trya and see sometime tomorrow if it 'picks up', but those Euro folks posting masses of small size headers (somewhere around 150-250KB per part) have flooded many newsgroups to the point of making it extremely difficult to 'scan' headers for anything in particular (sometimes hitting over 1million headers every couple of hours on certain groups).
Last night around 2200hrs (10pm Pacific time, as they are HQ'ed in LA/California) they stopped adding to various newsgroups and continued that for around 8+ hours. Various peers that are bound 'tight' to GN (like Blocknews) fairly quickly recovered to other peers, but some (like Binsearch as a good example) never did, loosing those 8+ hours into the 'bit bucket'.
As usual, I'll bet 99.999+% of the GN (or others) never noticed, but I did (as I use GN to scan groups for 'things', and it being Friday night in particular, 'Real Time' is a 'must get' and I noticed the problem right off). Like I said, Blocknews recovered pretty quickly (in times past, they always seem to do so whenever any of the other majors like GN or Astra run a bit of the rails, shows really good programming and operation there).
Any rate, around 06:30 (630am) this morning, it appeared that GN picked things back up. In the past when things have gone 'south' with GN, a few hours later they've 'backfilled' the missing headers in those groups in a few hours, but as of now (~1630, again Pacific time) they haven't.
I'll continue to monitor through this evening, but it just goes to show that even the 'high priced spread' does occasionally have problems. FYI, the usual 'suspect' in 'bad ops' (Astra) had zero problems all throughout this 'fade' at GN, and that 'best of the fills' Blocknews, really showed that it saw the problem and picked things up virtually right away.
UPDATE: As of 2200hrs (10pm) tonight, about 24hrs past the time when the 'event' started, still no 'backfills' to GN-US OR GN-EU. I'll trya and see sometime tomorrow if it 'picks up', but those Euro folks posting masses of small size headers (somewhere around 150-250KB per part) have flooded many newsgroups to the point of making it extremely difficult to 'scan' headers for anything in particular (sometimes hitting over 1million headers every couple of hours on certain groups).