Lines Matching full:traces

11 traces can be queried interactively, plots made from those results etc.
20 Because of this, what we actually want is to be able to query many traces
28 with the Python API to load multiple traces and query them in parallel.
31   traces. Over time, there would likely be a proliferation of slightly modified
33 * While the basics of parallelising queries on multiple traces on a single
44 While we've discussed querying traces so far, the experience for loading traces
45 from different traces should be just as good. This has historically been a big
50 knowing where traces live on the network filesystem and the directory layout.
53 version) that traces should match should match and traces meeting these criteria
57 * Interactively query ~1000+ traces in O(s) (for simple queries)
59 * Load traces from many sources with minimal ceremony. This should  include
79 are an elegant solution to the problem of loading traces from a diverse range
81 to describe both the protocol (i.e. the source) from which traces should be
82 fetched and the arguments (i.e. query parameters) which the traces should match.
95 Along with the job of yielding a list of traces, resolvers should also be
99 "flattening" results across many traces as discussed below.
101 ### Persisting loaded traces
102 Optimizing the loading of traces is critical for the O(s) query performance
103 we want out of batch trace processor. Traces are often accessed
105 Traces also take at least a few seconds to parse, eating up the budget for
108 To address this issue, we take the decision to keep all traces fully loaded in
112 For the moment, we restrict the loading and querying of traces to a
113 single machine. While querying n traces is "embarassngly parallel" and shards
119 The naive way to return the result of querying n traces is a list
126 a single table, containing the results from all the traces. However,
132 So suppose we query three traces with:
159 thread pools for loading traces and running queries. The parallelism and memory
166 constrained by a single machine and so can only load O(1000) traces.
167 For rare problems, there might only be a handful of traces matching a given
172 operating on a O(1000) traces with O(s) performance. Once the queries are
175 traces loading O(n cpus) traces at any one time.
183 across >1 machine. When querying traces, each trace is entirely independent of