diff for duplicates of <ef84de89c2597793d4cca5eee446ba90@localhost> diff --git a/a/1.txt b/N1/1.txt index e49debc..8af23d6 100644 --- a/a/1.txt +++ b/N1/1.txt @@ -1,25 +1,48 @@ On Thu, 19 May 2011 08:52:10 +0200, Alexander Zimmermann wrote: + + >> #define TCP_RTO_MAX ((unsigned)(120*HZ)) + >> #define TCP_RTO_MIN ((unsigned)(HZ/5)) + >> + >> So we're talking about a [200ms ; 120s] range no matter what. + > + > Why is 200ms a valid lower bound for initRTO? I'm aware of + > measurements that 1s is save for Internet, but I don't know of any + > studies that 200ms is save... + + TCP_RTO_MAX and TCP_RTO_MIN is the lower/upper bound for the RTO in + general, not for the initial RTO. RFC 2988 specify a lower bound of 1 + second but all operating system choose a lower one because at the time + where RFC 2988 was written the clock granularity was not that accurate. The + minimum RTO for FreeBSD is even 30ms! Furthermore, analysis had + demonstrated that a minimum RTO of 1 second badly breaks throughput in + environments faster then 33kB with minor packet loss rate (e.g. 1%). + + So yes, it CAN be wise to choose other lower/upper bounds. But keep in + mind that we should NOT artificial limit ourself. I can image data center + scenarios where a initial RTO of <1 match perfectly. + + Hagen diff --git a/a/content_digest b/N1/content_digest index 49f4038..8acb310 100644 --- a/a/content_digest +++ b/N1/content_digest @@ -27,27 +27,50 @@ "\n" "On Thu, 19 May 2011 08:52:10 +0200, Alexander Zimmermann wrote:\n" "\n" + "\n" + "\n" ">> #define TCP_RTO_MAX ((unsigned)(120*HZ))\n" + "\n" ">> #define TCP_RTO_MIN ((unsigned)(HZ/5))\n" + "\n" ">> \n" + "\n" ">> So we're talking about a [200ms ; 120s] range no matter what.\n" + "\n" "> \n" + "\n" "> Why is 200ms a valid lower bound for initRTO? I'm aware of\n" + "\n" "> measurements that 1s is save for Internet, but I don't know of any\n" + "\n" "> studies that 200ms is save... \n" "\n" + "\n" + "\n" "TCP_RTO_MAX and TCP_RTO_MIN is the lower/upper bound for the RTO in\n" + "\n" "general, not for the initial RTO. RFC 2988 specify a lower bound of 1\n" + "\n" "second but all operating system choose a lower one because at the time\n" + "\n" "where RFC 2988 was written the clock granularity was not that accurate. The\n" + "\n" "minimum RTO for FreeBSD is even 30ms! Furthermore, analysis had\n" + "\n" "demonstrated that a minimum RTO of 1 second badly breaks throughput in\n" + "\n" "environments faster then 33kB with minor packet loss rate (e.g. 1%).\n" "\n" + "\n" + "\n" "So yes, it CAN be wise to choose other lower/upper bounds. But keep in\n" + "\n" "mind that we should NOT artificial limit ourself. I can image data center\n" + "\n" "scenarios where a initial RTO of <1 match perfectly.\n" "\n" + "\n" + "\n" Hagen -a7a3b36fe06d3beebcef221c1485283f15f4580b8b150bd2c2ece34e741eb3c4 +27dc8c99b3a4ee120858a02a6029a5561a01ba407debb0922eeea6fe18285da5
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.