| 1 | |
|---|
| 2 | LEGEND f: fixed - c: changed - a: added - r: removed |
|---|
| 3 | |
|---|
| 4 | 1.1: |
|---|
| 5 | |
|---|
| 6 | web) |
|---|
| 7 | |
|---|
| 8 | c: rewritten short versus FQDN hostname detection: now works properly |
|---|
| 9 | with ganglia hosts not using FQDN hostnames |
|---|
| 10 | f: display of xml parsetime for overview. no longer display parsetime |
|---|
| 11 | for archive (no parsing done) |
|---|
| 12 | f: down/offline nodes are now properly marked in cluster image again |
|---|
| 13 | f: bug where "Unavailable" row would not be shown in overview summary |
|---|
| 14 | table |
|---|
| 15 | |
|---|
| 16 | packaging) |
|---|
| 17 | |
|---|
| 18 | c: completely redone and rewritten by Olivier Lahaye - thanks! |
|---|
| 19 | |
|---|
| 20 | jobmond) |
|---|
| 21 | |
|---|
| 22 | a: now supports SLURM Workload Manager! |
|---|
| 23 | a: warning if connecting to remote BATCH_SERVER is not supported by |
|---|
| 24 | selected BATCH_API |
|---|
| 25 | f: bug where incorrect commandline option would trigger traceback in |
|---|
| 26 | usage() |
|---|
| 27 | |
|---|
| 28 | jobarchived) |
|---|
| 29 | |
|---|
| 30 | a: now performs regular database Housekeeping every 20 job XML |
|---|
| 31 | iterations (previously only once at startup) |
|---|
| 32 | a: now checks if ARCHIVE_DATASOURCES are present in gmetad.conf |
|---|
| 33 | f: prevent an Exception to occur when determining datasource polling |
|---|
| 34 | interval |
|---|
| 35 | f: bug where config file handle was not closed |
|---|
| 36 | |
|---|
| 37 | 1.0: |
|---|
| 38 | |
|---|
| 39 | jobmond) |
|---|
| 40 | |
|---|
| 41 | a: now supports multiple udp send channels |
|---|
| 42 | a: now supports job arrays |
|---|
| 43 | c: updated Gmetric XDR protocol to version 3.1+ compatible |
|---|
| 44 | |
|---|
| 45 | c: gmond.conf parsing has been rewritten to handle include's and |
|---|
| 46 | multiple send channels |
|---|
| 47 | c: METRIC_MAX_VAL_LEN is now determined from gmond.conf |
|---|
| 48 | c: utilize new job monarch protocol |
|---|
| 49 | |
|---|
| 50 | f: can now handle new PBSQuery / pbs_python versions |
|---|
| 51 | f: default gmond.conf search location is now /etc/ganglia/gmond.conf |
|---|
| 52 | f: fatal error's are now printed to shell upon startup, not just syslog |
|---|
| 53 | f: more error checking and miscellanious bugfixes |
|---|
| 54 | |
|---|
| 55 | jobarchived) |
|---|
| 56 | |
|---|
| 57 | r: no longer use pyPgSQL for postgres database |
|---|
| 58 | c: now use psycopg2 module for postgres database |
|---|
| 59 | |
|---|
| 60 | a: job thread now utilizes db commits and rollbacks |
|---|
| 61 | a: now use USER/PASS authentication to database (in stead of hostbased) |
|---|
| 62 | |
|---|
| 63 | c: database schema: changed job_id to varchar to support job arrays |
|---|
| 64 | c: database schema: changed job_name max length to 255, just like |
|---|
| 65 | torque |
|---|
| 66 | c: database schema: added username/password role authentication |
|---|
| 67 | c: utilize new job monarch protocol |
|---|
| 68 | |
|---|
| 69 | f: job thread no longer hangs when insert/update of a job in database |
|---|
| 70 | fails |
|---|
| 71 | f: rewrite of job (finished) detection: all finished jobs again |
|---|
| 72 | properly detected |
|---|
| 73 | f: job checking now done post-parsing not while parsing |
|---|
| 74 | f: more error checking and miscellanious bugfixes |
|---|
| 75 | |
|---|
| 76 | web) |
|---|
| 77 | |
|---|
| 78 | r: removed Pie chart |
|---|
| 79 | r: removed TemplatePower |
|---|
| 80 | r: removed php ini_set's and time limit directive: should be handled in |
|---|
| 81 | php.ini |
|---|
| 82 | r: removed "Get Fresh Data" button: served no purpose anymore |
|---|
| 83 | a: now utilize Dwoo templates for html output |
|---|
| 84 | |
|---|
| 85 | a: now use USER/PASS authentication to database (in stead of hostbased) |
|---|
| 86 | a: ClusterImage now drops a shadow below nodes |
|---|
| 87 | a: RRDs now show "Last: Min: Avg: Max:" values in legend |
|---|
| 88 | |
|---|
| 89 | c: utilize new job monarch protocol |
|---|
| 90 | c: all templates rewritten from TemplatePower to Dwoo |
|---|
| 91 | c: graph.php now used for overview and archive |
|---|
| 92 | c: RRDs job start/finish line is now dashed green/red line with legend |
|---|
| 93 | |
|---|
| 94 | f: some dbase fields are now CAST to INT for php since postgres now |
|---|
| 95 | requires explicit casts |
|---|
| 96 | f: sort order descending/ascending is now correct |
|---|
| 97 | f: many, many speed and memory improvements |
|---|
| 98 | f: more error checking and miscellanious bugfixes |
|---|
| 99 | |
|---|
| 100 | 0.4: |
|---|
| 101 | |
|---|
| 102 | jobmond) |
|---|
| 103 | a: SGE support |
|---|
| 104 | thanks to: Dave Love - d(d.o.t)love(a.t)liverpool(d.o.t)ac(d.o.t)uk |
|---|
| 105 | for writing it! |
|---|
| 106 | a: LSF support |
|---|
| 107 | thanks to: Mahmoud Hanafi - mhanafi(a.t)csc(d.o.t)com |
|---|
| 108 | for writing it! |
|---|
| 109 | a: GMETRIC_TARGET is now parsed from gmond.conf |
|---|
| 110 | a: GMETRIC_BINARY is now looked for in PATH |
|---|
| 111 | f: queue selection support is now working |
|---|
| 112 | thanks to: Craig West - cwest(a.t)astro(d.o.t)umass(d.o.t)edu |
|---|
| 113 | for the patch |
|---|
| 114 | web) |
|---|
| 115 | a: large graphs link for job report |
|---|
| 116 | thanks to: Craig West - cwest(a.t)astro(d.o.t)umass(d.o.t)edu |
|---|
| 117 | a: SHOW_EMPTY_COLUMN, SHOW_EMPTY_ROW options for ClusterImage hostname parsing |
|---|
| 118 | |
|---|
| 119 | 0.3.1: |
|---|
| 120 | |
|---|
| 121 | other) |
|---|
| 122 | f: updated INSTALL since "addons" directory is not included by default anymore in Ganglia |
|---|
| 123 | thanks to: Steven DuChene linux(d.a.s.h)clusters(a.t)mindspring(d.o.t)com |
|---|
| 124 | for reporting it |
|---|
| 125 | |
|---|
| 126 | rpm) |
|---|
| 127 | f: add "addons" directory since it's not included by default anymore in Ganglia |
|---|
| 128 | f: properly rewrite WEBDIR path in %files when rebuilding rpms with Makefile |
|---|
| 129 | |
|---|
| 130 | web) |
|---|
| 131 | f: typo in empty_cpu variable: causing incorrect 'free cpu' count and similar errors |
|---|
| 132 | thanks to: Craig West - cwest(a.t)astro(d.o.t)umass(d.o.t)edu |
|---|
| 133 | for reporting it |
|---|
| 134 | f: changed erroneous domain detection a little |
|---|
| 135 | thanks to: Craig West - cwest(a.t)astro(d.o.t)umass(d.o.t)edu |
|---|
| 136 | for reporting it |
|---|
| 137 | a: now properly detects whether or not to use FQDN or short hostnames w/o domain |
|---|
| 138 | thanks to: Craig West - cwest(a.t)astro(d.o.t)umass(d.o.t)edu |
|---|
| 139 | thanks to: Jeffrey Sarlo - JSarlo(a.t)Central(d.o.t)UH(d.o.t)EDU |
|---|
| 140 | for the many testing and reporting it |
|---|
| 141 | |
|---|
| 142 | SPECIAL THANKS to the University of Houston for sending me a shirt! |
|---|
| 143 | |
|---|
| 144 | jobarchived) |
|---|
| 145 | f: properly catch postgres exception |
|---|
| 146 | f: don't use debug_message while loading config file |
|---|
| 147 | |
|---|
| 148 | 0.3: |
|---|
| 149 | |
|---|
| 150 | web) |
|---|
| 151 | a: allow per-cluster settings/override options: see CLUSTER_CONFS option |
|---|
| 152 | a: clusterimage can now draw nodes at x,y position parsed from hostname |
|---|
| 153 | see SORTBY_HOSTNAME for this in clusterconf/example.php |
|---|
| 154 | a: clusterimage nodes are now clickable: has link to all jobs from that host |
|---|
| 155 | a: clusterimage nodes now have a tooltip: displays hostname and jobids for now |
|---|
| 156 | a: jobmonarch logo image |
|---|
| 157 | thank to: Robin Day |
|---|
| 158 | for the design |
|---|
| 159 | a: rrd graph of running/queued jobs to overview |
|---|
| 160 | a: per-cluster settings for archive database |
|---|
| 161 | thanks to: Alexis Michon - alexis(d.o.t)michon(a.t)ibcp(d.o.t)fr |
|---|
| 162 | for the patch |
|---|
| 163 | |
|---|
| 164 | c: host archive view is now more complete and detailed in the same manner as |
|---|
| 165 | Ganglia's own host view |
|---|
| 166 | c: host archive view available metric list is now compiled from disk, |
|---|
| 167 | so that the detailed archive host view works even when the node is currently down. |
|---|
| 168 | c: removed size restrictions from detailed host archive view |
|---|
| 169 | |
|---|
| 170 | f: compatibility: removed php5 call |
|---|
| 171 | thanks to: Alexis Michon - alexis(d.o.t)michon(a.t)ibcp(d.o.t)fr |
|---|
| 172 | for the patch |
|---|
| 173 | f: prevent negative cpu/node calculation |
|---|
| 174 | thanks to: aloga(a.t)ifca(d.o.t)unican(d.o.t)es |
|---|
| 175 | for the patch |
|---|
| 176 | f: archive search not properly resetting nodes list |
|---|
| 177 | thanks to: Alexis Michon - alexis(d.o.t)michon(a.t)ibcp(d.o.t)fr |
|---|
| 178 | for the patch |
|---|
| 179 | f: detailed host view from jobarchive was broken since hostbased support of 0.2 |
|---|
| 180 | now host view is properly set and parsed again |
|---|
| 181 | thanks to: Alexis Michon - alexis(d.o.t)michon(a.t)ibcp(d.o.t)fr |
|---|
| 182 | for reporting the bug and suggesting a patch |
|---|
| 183 | f: bug where jobstart redline indicator in host detail graphs was set incorrectly |
|---|
| 184 | or not at all due to a miscalculation in job times |
|---|
| 185 | f: bug where hostimage headertext xoffset was miscalculated, causing the column names |
|---|
| 186 | to overlap their position when the columnname was longer than the columnvalues |
|---|
| 187 | |
|---|
| 188 | jobmond) |
|---|
| 189 | |
|---|
| 190 | a: syslog support |
|---|
| 191 | a: report number of running/queued jobs as seperate metrics |
|---|
| 192 | a: native gmetric support, much faster and cleaner! |
|---|
| 193 | thanks to: Nick Galbreath - nickg(a.t)modp(d.o.t)com |
|---|
| 194 | for writing it and allowing inclusion in jobmond |
|---|
| 195 | |
|---|
| 196 | f: crashing jobmond when multiple nodes amounts are requested in |
|---|
| 197 | a queued job: numeric_node variable not initialized properly |
|---|
| 198 | thanks to: aloga(a.t)ifca(d.o.t)unican(d.o.t)es |
|---|
| 199 | for supplying the patch |
|---|
| 200 | and many others for reporting and helping debug this |
|---|
| 201 | f: hanging/blocked, increased cpu usage and halted reporting |
|---|
| 202 | thanks to: Bas van der Vlies - basv(a.t)sara(d.o.t)nl |
|---|
| 203 | for discovering the origin of the bug |
|---|
| 204 | thanks to: Mickael Gastineau - gastineau(a.t)imcce(d.o.t)fr |
|---|
| 205 | for reporting it and testing the fix |
|---|
| 206 | thanks to: Craig West - cwest(a.t)astro(d.o.t)umass(d.o.t)edu |
|---|
| 207 | for reporting it and testing the fix |
|---|
| 208 | f: uninitialized variable in checkGmetricVersion() |
|---|
| 209 | thanks to: Peter Kruse - pk(a.t)q-leap(d.o.t)com |
|---|
| 210 | for the patch |
|---|
| 211 | f: undefined PBSError |
|---|
| 212 | thanks to: Peter Kruse - pk(a.t)q-leap(d.o.t)com |
|---|
| 213 | for reporting it |
|---|
| 214 | |
|---|
| 215 | r: SGE support broken |
|---|
| 216 | |
|---|
| 217 | jobarchived) |
|---|
| 218 | |
|---|
| 219 | a: can now use py-rrdtool api instead of pipes, much faster! |
|---|
| 220 | install py-rrdtool to use this |
|---|
| 221 | backwards compatible fails back to pipes if module not installed |
|---|
| 222 | |
|---|
| 223 | c: all XML input was uniencoded, which could cause errors, |
|---|
| 224 | now all properly converted to normal strings |
|---|
| 225 | |
|---|
| 226 | f: when XML data source (gmetad) is unavailable parsethread didn't return correctly |
|---|
| 227 | which caused a large number of threads to spawn while consuming large amounts of memory |
|---|
| 228 | f: autocreate clusterdirs in archivedir |
|---|
| 229 | f: unhandled gather exception |
|---|
| 230 | f: incorrect stop_timestamping when jobs finished |
|---|
| 231 | thanks to: Alexis Michon - alexis(d.o.t)michon(a.t)ibcp(d.o.t)fr |
|---|
| 232 | for finding and debugging/testing it |
|---|
| 233 | |
|---|
| 234 | 0.2: |
|---|
| 235 | |
|---|
| 236 | web) |
|---|
| 237 | f: misc. optimization and bugfixes |
|---|
| 238 | f: now fully compatible with latest PHP5 and PHP4 |
|---|
| 239 | |
|---|
| 240 | c: cluster image now incorporates small text descr. |
|---|
| 241 | c: monarch (cluster/host) images no longer displayed |
|---|
| 242 | for clusters that are not jobmond enabled |
|---|
| 243 | c: pie chart percentages are now cpu-based instead of node-based |
|---|
| 244 | |
|---|
| 245 | a: host template for Ganglia |
|---|
| 246 | adds a extra monarch host image to Ganglia's host overview |
|---|
| 247 | which displays/links to the jobs on that host |
|---|
| 248 | NOTE!: be sure to copy/install new template from addons/templates |
|---|
| 249 | a: (optional) nodes hostnames column |
|---|
| 250 | thanks to: Daniel Barthel - daniel(d.o.t)barthel(a.t)nottingham(d.o.t)ac(d.o.t)uk |
|---|
| 251 | for the suggestion |
|---|
| 252 | |
|---|
| 253 | jobmond) |
|---|
| 254 | |
|---|
| 255 | f: when a job metric is longer than maximum metric length, |
|---|
| 256 | the info is split up amongst multiple metrics |
|---|
| 257 | f: no longer exit when batch server is unavailable |
|---|
| 258 | thanks to: Peter Kruse - pk(a.t)q-leap(d.o.t)com |
|---|
| 259 | for the patch |
|---|
| 260 | f: fd closure bug causing stderr/stdout to remain open after daemonizing |
|---|
| 261 | |
|---|
| 262 | c: rearranged code to allow support for other batch systems |
|---|
| 263 | |
|---|
| 264 | a: (experimental) SGE (Sun Grid Engine) support as batch server |
|---|
| 265 | thanks to: Babu Sundaram - babu(a.t)cs(d.o.t)uh(d.o.t)edu |
|---|
| 266 | who developed it for a OSCAR's Google-SoC project |
|---|
| 267 | a: pidfile support |
|---|
| 268 | thanks to: Michael Jeanson - michael(a.t)ccs(d.o.t)usherbrooke(d.o.t)ca |
|---|
| 269 | for the patch |
|---|
| 270 | a: usage display |
|---|
| 271 | thanks to: Michael Jeanson - michael(a.t)ccs(d.o.t)usherbrooke(d.o.t)ca |
|---|
| 272 | for the patch |
|---|
| 273 | a: queue selection support: ability to specify which QUEUE's to get jobinfo from |
|---|
| 274 | thanks to: Michael Jeanson - michael(a.t)ccs(d.o.t)usherbrooke(d.o.t)ca |
|---|
| 275 | for the patch |
|---|
| 276 | |
|---|
| 277 | jobarchived) |
|---|
| 278 | |
|---|
| 279 | f: XML retrieval for Ganglia version >= 3.0.3 working properly again |
|---|
| 280 | f: database storing for Ganglia version >= 3.0.3 working properly again |
|---|
| 281 | f: fd closure bug causing stderr/stdout to remain open after daemonizing |
|---|
| 282 | |
|---|
| 283 | c: misc. bugfixes to optimize XML connections |
|---|
| 284 | c: misc. bugfixes for misc. minor issues |
|---|
| 285 | |
|---|
| 286 | a: cleaning of stale jobs in dbase: see JOB_TIMEOUT option |
|---|
| 287 | |
|---|
| 288 | 0.1.1: |
|---|
| 289 | |
|---|
| 290 | web) |
|---|
| 291 | |
|---|
| 292 | f: misc. layout bugs for overview & search |
|---|
| 293 | f: bug that occured when calculating the number of nodes when there |
|---|
| 294 | was more than one job running on a machine |
|---|
| 295 | |
|---|
| 296 | c: column requested memory is now optional through conf.php |
|---|
| 297 | c: search and overview tables are now full screen (100%) |
|---|
| 298 | c: overview jobnames are now cutoff at max 9 characters |
|---|
| 299 | to prevent (layout) scews in the tables |
|---|
| 300 | c: overview graphs are no longer downsized |
|---|
| 301 | |
|---|
| 302 | a: (optional) column 'queued' (since) in overview |
|---|
| 303 | a: search results (can) now have a SEARCH_RESULT_LIMIT |
|---|
| 304 | this increases performance of the query's significantly! |
|---|
| 305 | a: date/time format as displayed is now configurable through conf.php |
|---|
| 306 | |
|---|
| 307 | jobmond) |
|---|
| 308 | |
|---|
| 309 | a: now reports 'queued since' (or creation time) of jobs |
|---|
| 310 | |
|---|
| 311 | documentation) |
|---|
| 312 | |
|---|
| 313 | f: wrong e-mail adress in INSTALL (doh!) |
|---|
| 314 | |
|---|
| 315 | 0.1: |
|---|
| 316 | |
|---|
| 317 | - First public release |
|---|
