- To: news@sprint.net, news-admin@sprint.net
- Subject: eskinews.eskimo.com
- From: Robert Dinse <nanook@eskimo.com>
- Date: Sat, 17 Jul 1999 19:34:31 -0700 (PDT)
- cc: outages-list@eskimo.com
- Resent-Date: Sat, 17 Jul 1999 19:34:36 -0700
- Resent-From: outages-list@eskimo.com
- Resent-Message-ID: <"0Tb2q3.0.Th.yqJat"@mx1>
- Resent-Sender: outages-list-request@eskimo.com
- Prev by Date: News, DSL, and Meeting...
- Next by Date: Seattle Megapop problems
- Prev by thread: Seattle Megapop problems
- Next by thread: News, DSL, and Meeting...
- Index(es):
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
eskinews.eskimo.com
This is concerning my news feed from Sprint, which in theory is supposed
to be a full feed, but in practice hasn't been for at least the last six
months.
I replaced my news server hoping to get something closer to a full feed;
and it helped, to a point.
At 4am, I get something resembling a full feed, between
10-15mbytes/minute, but at 7:15pm, the news slows to around 3-4mb/minute.
With this new server I think I have pretty much eliminated load on my
machine being a factor, it's averaging 11% CPU occupancy:
7:20pm up 15:36, 4 users, load average: 0.53, 0.64, 0.55
procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
2 0 0 1640 17100 95104 46704 0 0 20 39 64 82 5 6 89
The load factor of .53 is on a 4 CPU machine, which means there are more
than three idle CPU's ready to do something most of the time.
This thing has a 10-base-T segment all by itself, to a HUB on a FDDI LAN,
the FDDI occupancy is 3%, T1 link occupancy is about 50% at present:
== PATH statistics ============= 1========= 3======== 3B======== 3C========= 4=
Rcvd Packets 39419 19110 995 - 18848
Bytes 14476449 4266429 123266 - 4159910
Err: CRC 0 0 0 - 0
Framing 0 0 0 - 0
Too Long 0 0 0 - 0
Lost 0 0 0 - 0
Xmit Packets 38235 19478 1054 - 19484
Bytes 8995839 6842799 236459 - 6734520
Err: Deferred 0 0 0 - 0
Collision 0 0 0 - 0
Late Collisions 0 0 0 - 0
Xcess Collision 0 0 0 - 0
Carrier Loss 0 0 0 - 0
Underrun 0 0 0 - 0
Discard: Buf Overrun 0 0 0 - 0
Congestion 0 0 0 - 0
Utilization: (%) 3 48 1 - 47
Rcv Good : Pkt/Sec 656 318 16 - 314
The network connectivity doesn't look too bad:
--- news-central.sprintlink.net ping statistics ---
101 packets transmitted, 101 packets received, 0% packet loss
round-trip min/avg/max = 46.5/58.3/135.9 ms
Average latency could be better, but 58.3ms isn't horrid.
So what I am trying to understand is why news slows to a crawl in the day
even when the load on my machine is nil, and the network load is not
particularly high.