| To: | Nathan Scott <nathans@xxxxxxxxxx> |
|---|---|
| Subject: | Re: pcp updates: merges, docs, qa |
| From: | fche@xxxxxxxxxx (Frank Ch. Eigler) |
| Date: | Fri, 30 Oct 2015 11:17:30 -0400 |
| Cc: | PCP <pcp@xxxxxxxxxxx> |
| Delivered-to: | pcp@xxxxxxxxxxx |
| In-reply-to: | <1674933679.62705114.1445899319680.JavaMail.zimbra@xxxxxxxxxx> (Nathan Scott's message of "Mon, 26 Oct 2015 18:41:59 -0400 (EDT)") |
| References: | <1674933679.62705114.1445899319680.JavaMail.zimbra@xxxxxxxxxx> |
| User-agent: | Gnus/5.1008 (Gnus v5.10.8) Emacs/21.4 (gnu/linux) |
> [...] > pmcd: implement dynamic switching of monitored containers > [...] How does commit work when composed with shared libpcp connections? It seems as though if an application does two vanilla c1 = pmNewContext(host) c2 = pmNewContext(host) calls, it will get a shared socket - even after your libpcp sharing fixes, since their 'c_flags' == 0. Thereafter, will pmStore(c1, "container=foo") affect c2? A more substantial test case for this, which would start exercising the connection sharing fix, would be welcome. - FChE |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | [Bug 1275293] Suboptimal client initialization with -L, bugzilla |
|---|---|
| Next by Date: | Re: Introducing pmrep - Performance Metrics Reporter, Frank Ch. Eigler |
| Previous by Thread: | pcp updates: merges, docs, qa, Nathan Scott |
| Next by Thread: | pcp updates: trivia, Ken McDonell |
| Indexes: | [Date] [Thread] [Top] [All Lists] |