This document describes the changes made to the Observer application.
The time needed for loading a crashump into the crashdump viewer would earlier grow exponentially with the size of the crashdump file. Reading a file of 20M would take a couple of minutes, and for a dump of 250M it would take between 1 and 2 hours. This has been solved.
Earlier, all processes, timers, funs or ets-tables would be loaded into the memory of the crashdump viewer node before sending it on to the web server. This has been changed and the pages are now sent to the web server in chunks.
A security function in newer web browsers prevents a full file path to be sent from an HTML file input field, i.e. the field needed to implement the "Browse" button when loading a file into the crashdump viewer. To overcome this, the file input field is no longer used. Instead a normal text input field is used, and the user needs to manually insert the complete file path. For convenience, a shell script and a batch file are added to the observer application. These can be used to start the crashdump_viewer and a browser and load a file - with the file name given from the command line. The shell script and batch file are called cdv and cdv.bat respectively, and can be found in the priv dir of the observer application.
Own Id: OTP-9051 Aux Id: seq11789
The multitrace.erl installation example file is now installed in the examples directory. (Thanks to Peter Lemenkov.)
Own Id: OTP-8857
The test suite has been updated for R14A.
Own Id: OTP-8708
Misc updates
Own Id: OTP-8456
Major improvements of the Erlang distribution for Erlang runtime systems with SMP support. Previously distribution port locks were heavily contended, and all encoding and decoding for a specific distribution channel had to be done in sequence. Lock contention due to the distribution is now negligible and both encoding and decoding of Erlang messages can be done in parallel.
The old atom cache protocol used by the Erlang distribution has been dropped since it effectively prevented all parallel encoding and decoding of messages passed over the same distribution channel.
A new atom cache protocol has been introduced which isolates atom cache accesses and makes parallel encoding and decoding of messages passed over the same distribution channel possible. The new atom cache protocol also use an atom cache size 8 times larger than before. The new atom cache protocol is documented in the ERTS users guide.
Erlang messages received via the distribution are now decoded by the receiving Erlang processes without holding any distribution channel specific locks. Erlang messages and signals sent over the distribution are as before encoded by the sending Erlang process, but now without holding any distribution channel specific locks during the encoding. That is, both encoding and decoding can be and are done in parallel regardless of distribution channel used.
The part that cannot be parallelized is the atom cache updates. Atom cache updates are therefore now scheduled on the distribution port. Since it is only one entity per distribution channel doing this work there is no lock contention due to the atom cache updates.
The new runtime system does not understand the old atom cache protocol. New and old runtime systems can however still communicate, but no atom cache will be used.
Own Id: OTP-7774
Own Id: OTP-7519
The copyright notices have been updated.
Own Id: OTP-7851
Minor Makefile changes.
Own Id: OTP-6689
Obsolete guard tests (such as list()) have been replaced with the modern guard tests (such as is_list()).
Own Id: OTP-6725
This application has been updated to eliminate warnings by Dialyzer.
Own Id: OTP-6551
Several minor bugs and race conditions eliminated in the runtime_tools and observer applications.
Own Id: OTP-6265
Crash dump with large integers could crash the
Own Id: OTP-6301
Fixed a bug in
Own Id: OTP-6075
The Observer application has been recompiled because of a compiler bug.
Own Id: OTP-5700
Own Id: OTP-5408