| To: | Ken McDonell <kenj@xxxxxxxxxxxxxxxx>, Lukas Berk <lberk@xxxxxxxxxx> |
|---|---|
| Subject: | Re: [pcp] problem in qa/884 (libpcp_web) |
| From: | Nathan Scott <nathans@xxxxxxxxxx> |
| Date: | Tue, 6 Sep 2016 03:36:24 -0400 (EDT) |
| Cc: | PCP <pcp@xxxxxxxxxxx> |
| Delivered-to: | pcp@xxxxxxxxxxx |
| In-reply-to: | <5257a34f-b310-15b3-d2e2-78cb35bd48de@xxxxxxxxxxxxxxxx> |
| References: | <5257a34f-b310-15b3-d2e2-78cb35bd48de@xxxxxxxxxxxxxxxx> |
| Reply-to: | Nathan Scott <nathans@xxxxxxxxxx> |
| Thread-index: | 0P9Sw8OEIJAL7Q916Acb7++UzEw5Wg== |
| Thread-topic: | problem in qa/884 (libpcp_web) |
----- Original Message ----- > I'm seeing problems with qa/884. > > The checked in 884.out contains suspicious negative 32-bit values for > osd/stat_bytes* ... > Hmm, yes, it looks like the parsing code is assuming signed 32 bit values for all JSON numbers. I think to fix this Lukas we'll need to extend the "flags" field in the json_metric_desc to specify type? (in a back-compat way such that no flag means signed int). cheers. -- Nathan |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Urgent Order, Luke Sales |
|---|---|
| Next by Date: | Re: [pcp] pmclusterd versus other solutions, Mark Goodwin |
| Previous by Thread: | problem in qa/884 (libpcp_web), Ken McDonell |
| Next by Thread: | Re: [pcp] problem in qa/884 (libpcp_web), Lukas Berk |
| Indexes: | [Date] [Thread] [Top] [All Lists] |