From 06/23/2016 to 07/22/2016
- 03:05 PM Feature #24 (New): Key-value store CAAPI: split data log and index log
- If the data log and index log are separate, it enables one to use preexisting logs for the CAAPI. Also, the same data...
- 02:55 PM Feature #23 (On Hold): Android support
- Long Le reports that using Java (and JNA) works easily enough on Android. It'd be nice to have the already *sort of* ...
- 09:49 AM Bug #17 (Feedback): Issue with gcl-create and log subscription
- I believe this was fixed with commit:555a2264 (Kaz found and fixed this). Metadata wasn't retained with the in-memor...
- 08:53 AM Bug #22: Darwin: Compiling ep fails: EVP_DigestSignFinal not found
- Mea culpa. Some time back I updated the ep Makefile so that it would be more like the others (in particular, using t...
- 08:29 AM Bug #22: Darwin: Compiling ep fails: EVP_DigestSignFinal not found
- Here's a diff that fixes it for me
diff --git a/ep/Makefile b/ep/Makefile
index 11b3626..7ee2e30 100644
- 08:27 AM Bug #22: Darwin: Compiling ep fails: EVP_DigestSignFinal not found
- I'm not sure how to reopen a bug in RedMine. The Status field shows "Closed" and has no way to edit it.
Not a bi...
- 07:32 AM Bug #22 (Closed): Darwin: Compiling ep fails: EVP_DigestSignFinal not found
- It turns out that LIBSEARCH was supposed to do this but, due to a cut-and-paste error, the Makefile was using -I inst...
- 06:42 AM Bug #22 (Closed): Darwin: Compiling ep fails: EVP_DigestSignFinal not found
- Under Darwin, running make in gdp/ep fails because /opt/local/lib/libcrypto.dylib is not used at link time
How to ...
- 04:07 PM Bug #14: GDP logging
- I *think* I found a race condition that would explain this, but I had trouble when I tried to deploy the fix. I'm te...
- 02:43 PM Bug #14 (In Progress): GDP logging
- Caused two days of data to be lost from sensors deployed in Berkeley. Eric currently working on this
- 03:50 PM Bug #20 (Closed): size_t vs ssize_t for evbuffer_copyout
- Changed it to return 0 instead of -1 on error. The error can only occur if you have called `evbuffer_freeze` on it, ...
- 02:02 AM Bug #20 (Closed): size_t vs ssize_t for evbuffer_copyout
- `evbuffer_copyout` returns a type of `ev_ssize_t`, which maps to `ssize_t` allowing for signed values. It returns *th...
- 02:41 PM Bug #19 (Closed): Error handling in web-visualization for sensor nodes
- partial fix for #21 fixes this as well. Closing for now.
- 02:39 PM Bug #21 (In Progress): "500: internal server error" in the visualization tool.
- 02:19 PM Bug #21: "500: internal server error" in the visualization tool.
- commit:32a0a36 fixes the issue "500: internal server error" and provides a more meaningful result, i.e. empty set on ...
- 01:42 PM Bug #21 (Closed): "500: internal server error" in the visualization tool.
- Example URL with the issue: http://gdp.cs.berkeley.edu:8888/datasource?logname=edu.berkeley.eecs.swarmlab.device.c098...
- 02:25 PM Bug #19 (Closed): Error handling in web-visualization for sensor nodes
- The web-visualization currently does not handle any erroneous situations from the GDP data sources (no-data, or any e...
- 09:50 AM Bug #17 (Closed): Issue with gcl-create and log subscription
- Filing on behalf of *Nima*:
I noticed that these seems to be an issue with gcl-create and log subscription. I'm ab...
- 09:26 AM Bug #16 (Closed): Threaded GDP Clients Fail
- On behalf of *Rico*:
I've been doing experiments for our "central" service that I've mentioned before. Currently, ...
- 09:22 AM Bug #15 (Closed): Assertion error in Python API
- From *Nima Mousavi*: *I'm facing some issues while trying to access GDP logs. I got the following error, but the issu...
- 04:08 PM Bug #14 (Closed): GDP logging
- The mqtt-gdp-gateway starts returning the follwoing error randomly after logging data for some time.
- 08:27 PM Feature #13 (In Progress): Flushing administrative output
- 08:27 PM Feature #13: Flushing administrative output
- I guess I had assumed that this was line buffered, but that was a bad assumption. Another point: it should also lock...
- 08:21 PM Feature #13 (Closed): Flushing administrative output
- There ought to be a `fflush(fp);` at source:gdplogd/logd_admin.c#L229
This ensures that all the data from a single c...
- 07:20 PM Bug #12 (Closed): Connection reset for clients
- *Summary:* followed by a `Connection reset by peer` error, everything fails. Router process is verified to be running...
- 01:45 PM Bug #11 (Feedback): No avahi package in homebrew (Mac OS X)
- I updated the documentation to make this clear and added a compile flag to make it easier to compile Zeroconf out. I...
- 12:58 PM Bug #10 (Feedback): Document / auto-configure the CRYPTOROOT option.
- Changed Makefiles to always search /usr/local and /opt/local, completely eliminating CRYPTOROOT. Fixed in commit:475...
- 11:35 AM Document: PCARI visit, July 11th 2016
- 09:41 AM Bug #5 (Closed): libs directory should not be in git; "make all" should create it
- Christopher has confirmed that this works on RedHat
- 03:10 PM Bug #11 (Closed): No avahi package in homebrew (Mac OS X)
- It seems that there's no avahi package in home brew. I didn't see a good way to install avahi from source, as it appe...
- 01:31 PM Bug #10 (Closed): Document / auto-configure the CRYPTOROOT option.
- Naively installing on Mac OS X using brew, I got errors that openssl couldn't be found:
- 12:50 PM Bug #9 (Closed): Router terminating connections abruptly
- Eric says:
[After creating a new log on a log-server connected to a co-located router, ] For some reason I don't s...
- 04:44 PM Document: The Cloud is Not Enough: Saving IoT from the Cloud
- Appears in HotCloud 2015.
- Hotcloud paper
- 04:43 PM Document: Toward a Global Data Infrastructure
- Appears in IEEE Internet Computing (Volume:20 , Issue: 3 ). **Not to be made public without a IEEE copyright statem...
- IEEE Internet Computing article
- 01:01 PM Bug #5 (Closed): libs directory should not be in git; "make all" should create it
- Today the `libs` directory has symbolic links checked into git. This creates problems for various bindings. The `li...
- 11:47 AM Feature #1 (Closed): Should multiread return an error for an empty log?
- As it stands, multiread on an empty log returns a 404 error (since it's trying to read a record that doesn't exist). ...
Also available in: Atom