Re: About the error "too many files opened"

[Danny Bickson]

Hi Jason!

It can definitely be a bug that not all open file descriptors are closed. In order that I will take a look, can you tell me which application are you using?

thanks

Dr. Danny Bickson

Project Scientist, Machine Learning Dept.

Carnegie Mellon University

On Thu, Dec 6, 2012 at 10:23 AM, Jason Chao <jaso...@gmail.com> wrote:

Hi,

I'm now handling a large bulk of data. When the data is preprocessing, the log shows 253 shards will be created. But during the phase 3, error message "Too many files opened" is showed.

The ulimit of the machine I'm using is 4096, and I can't get more. So, is 4096 not enough for generating and using 253 shards?

Or, are there any unused files not being closed during the preprocessing?

Thanks.

[Danny Bickson]

Can you send us also the exact command you used and the full output you got.

What is your data size? (send me the 3 line of the header files of the training, validation and test data).

Thanks

Dr. Danny Bickson

Project Scientist, Machine Learning Dept.

Carnegie Mellon University

On Thu, Dec 6, 2012 at 10:32 AM, Jason Chao <jaso...@gmail.com> wrote:

Hi, Danny,

I was just using als-tensor.

[Danny Bickson]

p.s.

Hi Jason, I fixed one case of opened file descriptor that was not closed in my code, but I suspect there is a more severe issue in the sharder or engine.

Please retake the latest (using hg pull; hg update) and recompile (using make clean; make cf). I will ask Aapo to take a deeper look into this.

Dr. Danny Bickson

Project Scientist, Machine Learning Dept.

Carnegie Mellon University

[Danny Bickson]

p.p.s

You can also try an limit the number of shards as a workaround using "nshards XX" command line flag, for example

./toolkits/collaborative_filtering/als_tensor --training=time_smallnetflix --max_iter=10 --minval=1 --maxval=5 nshards 3

Note that you have to delete all intermediate processed files before running this command using

rm time_smallnetflix.*

Also note the sepcial syntax of nshards command (without "--" and without "=")

Dr. Danny Bickson

Project Scientist, Machine Learning Dept.

Carnegie Mellon University

[Aapo Kyrola]

Hi Jason,

thanks for this report. 

Two issues:

1) There was a bug, that the file descriptors were not closed after sharding.

Please "hg pull", and you problem should be away.

2) the reason you are opening huge number of files is that you use "niothreads=5"

setting. This means, that for each file, the system opens 2 * 5 file descriptors (one for

read, one for write for each thread).  (This is also bad design, sicne I think only one

fd would be needed / thread).  Anyway, using so many I/O threads is not going to

help your performance, so I suggest you set niothreads=2 in your config file.

Danny is also right that the system opens the write descriptors even if it is not

using them in SGD and other apps.  I will fix this later, but you should be good to go :).

Thanks,

Aapo

On Dec 6, 2012, at 7:23 AM, Jason Chao wrote:

Yes. 

I just turned to try sgd, and set a larger mem-budget and the shards number reduced to 195. But the error still appeared.

The command was: ./toolkits/collaborative_filtering/sgd --training=ptype

The training data size is: 4870561 538535 932553596

The part of output:

INFO:     sharder.hpp(determine_number_of_shards:263): Determining number of shards automatically.

INFO:     sharder.hpp(determine_number_of_shards:266): Assuming available memory is 2500 megabytes. 

INFO:     sharder.hpp(determine_number_of_shards:267):  (This can be defined with configuration parameter 'membudget_mb')

INFO:     sharder.hpp(determine_number_of_shards:273): Determining maximum shard size: 312.5 MB.

INFO:     sharder.hpp(determine_number_of_shards:282): Number of shards to be created: 156

INFO:     sharder.hpp(execute_sharding:235): Max vertex id: 5409095

INFO:     sharder.hpp(start_phase:330): Starting phase: 1

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 5.52136e-08%

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 1.25956%

....

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 99.5053%

INFO:     sharder.hpp(end_phase:368): Ending phase: 1

INFO:     sharder.hpp(compute_partitionintervals:288): Number of shards: 156

INFO:     sharder.hpp(compute_partitionintervals:289): Edges per shard: 5089282

INFO:     sharder.hpp(compute_partitionintervals:290): Max vertex id: 5409095

INFO:     sharder.hpp(compute_partitionintervals:309): Interval: 0 - 4874364

INFO:     sharder.hpp(compute_partitionintervals:309): Interval: 4874365 - 4878297

...

INFO:     sharder.hpp(compute_partitionintervals:309): Interval: 5409096 - 5409096

INFO:     sharder.hpp(compute_partitionintervals:318): Computed intervals.

INFO:     sharder.hpp(execute_sharding:235): Max vertex id: 5409095

INFO:     sharder.hpp(start_phase:330): Starting phase: 2

DEBUG:    sharder.hpp(start_phase:350): Shoveling bufsize: 4201056

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 5.52136e-08%

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 1.25956%

...

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 98.2457%

DEBUG:    binary_adjacency_list.hpp(read_edges:133): 99.5053%

INFO:     sharder.hpp(end_phase:368): Ending phase: 2

INFO:     sharder.hpp(write_shards:442): Starting final processing for shard: 0

DEBUG:    sharder.hpp(write_shards:453): Shovel size:366699168 edges: 5093044

INFO:     sharder.hpp(write_shards:442): Starting final processing for shard: 1

DEBUG:    sharder.hpp(write_shards:453): Shovel size:366550416 edges: 5090978

...

INFO:     sharder.hpp(create_degree_file:605): (Degree proc.) Sub-window: [5403615 - 5407101]

INFO:     stripedio.hpp(open_session:335): Opened read-session: 465(2813) for ptype.edata_azv.153_156.adj

INFO:     sharder.hpp(create_degree_file:601): Interval: 5407102 5409095

INFO:     sharder.hpp(create_degree_file:605): (Degree proc.) Sub-window: [5407102 - 5409095]

INFO:     stripedio.hpp(open_session:335): Opened read-session: 466(2813) for ptype.edata_azv.154_156.adj

INFO:     sharder.hpp(create_degree_file:601): Interval: 5409096 5409095

 === REPORT FOR sharder() ===

[Numeric]

niothreads:             5

stripesize:             1.07374e+09

subwindow:              5e+06

[Timings]

blockload:              33.0899s         (count: 7589, min: 2e-06s, max: 0.608211, avg: 0.00436024s)

degrees.runtime:                61.0483 s

execute_sharding:               5141.67 s

memoryshard_create_edges:               22.6198s         (count: 155, min: 0.071653s, max: 0.963084, avg: 0.145934s)

preada_now:             39.1009s         (count: 7744, min: 1e-06s, max: 0.704726, avg: 0.00504918s)

preprocessing:          1501.35 s

read_next_vertices:             70.3317s         (count: 24025, min: 0s, max: 4.48643, avg: 0.00292744s)

stream_ahead:           58.2758s         (count: 155, min: 0.075499s, max: 32.0114, avg: 0.375973s)

stream_ahead.0:         32.0114 s

stream_ahead.1:         0.184422 s

stream_ahead.10:                0.169461 s

stream_ahead.98:                0.139486 s

stream_ahead.99:                0.118636 s

[Other]

app:    sharder

INFO:     ptype.cpp(convert_input:173): Successfully finished sharding for ptype

INFO:     ptype.cpp(convert_input:174): Created 156 shards.

INFO:     stripedio.hpp(io_thread_loop:662): Thread for multiplex :0 starting.

INFO:     stripedio.hpp(io_thread_loop:662): Thread for multiplex :0 starting.

INFO:     stripedio.hpp(io_thread_loop:662): Thread for multiplex :0 starting.

INFO:     stripedio.hpp(io_thread_loop:662): Thread for multiplex :0 starting.

INFO:     stripedio.hpp(io_thread_loop:662): Thread for multiplex :0 starting.

INFO:     graphchi_engine.hpp(graphchi_engine:147): Initializing graphchi_engine. This engine expects 64-byte edge data. 

INFO:     graphchi_engine.hpp(load_vertex_intervals:861): shard: 0 - 4874364

INFO:     graphchi_engine.hpp(load_vertex_intervals:861): shard: 4874365 - 4878297

...

INFO:     graphchi_engine.hpp(load_vertex_intervals:861): shard: 5407102 - 5409095

INFO:     graphchi_engine.hpp(load_vertex_intervals:861): shard: 5409095 - 5409095

INFO:     stripedio.hpp(open_session:333): Opened write-session: 0(2813) for ptype_degs.bin

INFO:     graphchi_engine.hpp(run:566): GraphChi starting

INFO:     graphchi_engine.hpp(run:567): Licensed under the Apache License 2.0

INFO:     graphchi_engine.hpp(run:568): Copyright Aapo Kyrola et al., Carnegie Mellon University (2012)

INFO:     stripedio.hpp(open_session:333): Opened write-session: 1(2825) for ptype.208B.vout

INFO:     stripedio.hpp(open_session:333): Opened write-session: 2(2837) for ptype.edata_azv.e64B.0_156

INFO:     stripedio.hpp(open_session:335): Opened read-session: 3(2848) for ptype.edata_azv.0_156.adj

INFO:     stripedio.hpp(open_session:333): Opened write-session: 4(2855) for ptype.edata_azv.e64B.1_156

INFO:     stripedio.hpp(open_session:335): Opened read-session: 5(2866) for ptype.edata_azv.1_156.adj

INFO:     stripedio.hpp(open_session:333): Opened write-session: 6(2873) for ptype.edata_azv.e64B.2_156

INFO:     stripedio.hpp(open_session:335): Opened read-session: 7(2884) for ptype.edata_azv.2_156.adj

...

INFO:     stripedio.hpp(open_session:335): Opened read-session: 135(4036) for ptype.edata_azv.66_156.adj

INFO:     stripedio.hpp(open_session:333): Opened write-session: 136(4043) for ptype.edata_azv.e64B.67_156

INFO:     stripedio.hpp(open_session:335): Opened read-session: 137(4054) for ptype.edata_azv.67_156.adj

INFO:     stripedio.hpp(open_session:333): Opened write-session: 138(4061) for ptype.edata_azv.e64B.68_156

INFO:     stripedio.hpp(open_session:335): Opened read-session: 139(4072) for ptype.edata_azv.68_156.adj

INFO:     stripedio.hpp(open_session:333): Opened write-session: 140(4079) for ptype.edata_azv.e64B.69_156

INFO:     stripedio.hpp(open_session:335): Opened read-session: 141(4090) for ptype.edata_azv.69_156.adj

ERROR:    stripedio.hpp(get_filesize:736): Could not open file ptype.edata_azv.e64B.70_156 error: Too many open files

sgd: ../../src/io/stripedio.hpp:737: size_t graphchi::get_filesize(std::string): Assertion `false' failed.

Aapo Kyrola
Ph.D. student, http://www.cs.cmu.edu/~akyrola

GraphChi: Big Data - small machine: http://graphchi.org

[Aapo Kyrola]

'Hg pull' was for getting the latest version from version control. As described, I fixed a bug in the code. You are using very old code: please take new one from GitHub: GitHub.com/graphchi 

Sent from my iPhone

On Jan 19, 2014, at 19:09, cxs...@gmail.com wrote:

Hi Aapo,

Thanks for helping us fix bugs so patiently.

I have a few questions.

What is 'hg pull'? How does it relate to closing file descriptors?

I tried the second solution which is giving a new parameter niothreads = 2 but it didn't work.

The command I used was:

xiao@Xiao-MacBookPro:~Desktop/graphchi-cpp-master$ bin/example_apps/communitydetection file user_relation filetype edgelist niothreads 2

The output I got was:

 === REPORT FOR sharder() ===

[Timings]
edata_flush:        10.028s     (count: 1397, min: 0.000297s, max: 1.39358, avg: 0.00717825s)
execute_sharding:        51.6636 s
finish_shard.sort:        12.7329s     (count: 15, min: 0.812712s, max: 0.924562, avg: 0.848858s)
preprocessing:        56.7182 s
shard_final:        31.3498s     (count: 15, min: 1.67133s, max: 3.91914, avg: 2.08999s)
[Other]
app:    sharder
INFO:     sharder.hpp(done:903): Created 15 shards, for 182575156 edgesSuccessfully finished sharding for rrall
INFO:     conversions.hpp(convert:651): Created 15 shards.
DEBUG:    stripedio.hpp(stripedio:271): Start io-manager with 2 threads.
INFO:     graphchi_engine.hpp(graphchi_engine:154): Initializing graphchi_engine. This engine expects 8-byte edge data.
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 0 - 48476
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 48477 - 82149
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 82150 - 131978
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 131979 - 195995
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 195996 - 240472
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 240473 - 281098
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 281099 - 332498
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 332499 - 396074
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 396075 - 477084
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 477085 - 570296
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 570297 - 667644
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 667645 - 787562
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 787563 - 901706
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 901707 - 1030931
INFO:     chifilenames.hpp(load_vertex_intervals:400): shard: 1030932 - 1192969
INFO:     graphchi_engine.hpp(run:737): GraphChi starting
INFO:     graphchi_engine.hpp(run:738): Licensed under the Apache License 2.0
INFO:     graphchi_engine.hpp(run:739): Copyright Aapo Kyrola et al., Carnegie Mellon University (2012)
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97374232, rrall.edata..Z.e8B.0_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 98048472, rrall.edata..Z.e8B.1_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97323000, rrall.edata..Z.e8B.2_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97321576, rrall.edata..Z.e8B.3_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97546320, rrall.edata..Z.e8B.4_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97299032, rrall.edata..Z.e8B.5_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97298976, rrall.edata..Z.e8B.6_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97299144, rrall.edata..Z.e8B.7_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97299032, rrall.edata..Z.e8B.8_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97299056, rrall.edata..Z.e8B.9_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97299696, rrall.edata..Z.e8B.10_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97299088, rrall.edata..Z.e8B.11_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97297888, rrall.edata..Z.e8B.12_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97298352, rrall.edata..Z.e8B.13_15sizeof(ET): 8
DEBUG:    slidingshard.hpp(sliding_shard:213): Total edge data size: 97297384, rrall.edata..Z.e8B.14_15sizeof(ET): 8
INFO:     graphchi_engine.hpp(print_config:132): Engine configuration:
INFO:     graphchi_engine.hpp(print_config:133):  exec_threads = 2
INFO:     graphchi_engine.hpp(print_config:134):  load_threads = 4
INFO:     graphchi_engine.hpp(print_config:135):  membudget_mb = 800
INFO:     graphchi_engine.hpp(print_config:136):  blocksize = 1048576
INFO:     graphchi_engine.hpp(print_config:137):  scheduler = 1
INFO:     graphchi_engine.hpp(run:773): Start iteration: 0
INFO:     graphchi_engine.hpp(run:852): 1.06965s: Starting: 0 -- 48476
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 0 - 48476
DEBUG:    graphchi_engine.hpp(run:880): Allocation 48477 vertices, sizeof:64 total:3102528
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.00438312 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 4.55519s: Starting: 48477 -- 82149
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 48477 - 82149
DEBUG:    graphchi_engine.hpp(run:880): Allocation 33673 vertices, sizeof:64 total:2155072
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.00613522 number of blocks: 94
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 7.95491s: Starting: 82150 -- 131978
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 82150 - 131978
DEBUG:    graphchi_engine.hpp(run:880): Allocation 49829 vertices, sizeof:64 total:3189056
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.00748464 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 12.0595s: Starting: 131979 -- 195995
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 131979 - 195995
DEBUG:    graphchi_engine.hpp(run:880): Allocation 64017 vertices, sizeof:64 total:4097088
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.00935553 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 17.3931s: Starting: 195996 -- 240472
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 195996 - 240472
DEBUG:    graphchi_engine.hpp(run:880): Allocation 44477 vertices, sizeof:64 total:2846528
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0116936 number of blocks: 94
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 22.0025s: Starting: 240473 -- 281098
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 240473 - 281098
DEBUG:    graphchi_engine.hpp(run:880): Allocation 40626 vertices, sizeof:64 total:2600064
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0133731 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 26.1992s: Starting: 281099 -- 332498
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 281099 - 332498
DEBUG:    graphchi_engine.hpp(run:880): Allocation 51400 vertices, sizeof:64 total:3289600
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0149012 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 30.7277s: Starting: 332499 -- 396074
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 332499 - 396074
DEBUG:    graphchi_engine.hpp(run:880): Allocation 63576 vertices, sizeof:64 total:4068864
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0166875 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 35.382s: Starting: 396075 -- 477084
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 396075 - 477084
DEBUG:    graphchi_engine.hpp(run:880): Allocation 81010 vertices, sizeof:64 total:5184640
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0190795 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 40.2552s: Starting: 477085 -- 570296
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 477085 - 570296
DEBUG:    graphchi_engine.hpp(run:880): Allocation 93212 vertices, sizeof:64 total:5965568
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0217838 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 45.5556s: Starting: 570297 -- 667644
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 570297 - 667644
DEBUG:    graphchi_engine.hpp(run:880): Allocation 97348 vertices, sizeof:64 total:6230272
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0247476 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 50.9517s: Starting: 667645 -- 787562
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 667645 - 787562
DEBUG:    graphchi_engine.hpp(run:880): Allocation 119918 vertices, sizeof:64 total:7674752
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0278123 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 56.7501s: Starting: 787563 -- 901706
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 787563 - 901706
DEBUG:    graphchi_engine.hpp(run:880): Allocation 114144 vertices, sizeof:64 total:7305216
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0318638 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 62.8523s: Starting: 901707 -- 1030931
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 901707 - 1030931
DEBUG:    graphchi_engine.hpp(run:880): Allocation 129225 vertices, sizeof:64 total:8270400
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.0348053 number of blocks: 93
INFO:     graphchi_engine.hpp(run:889): Start updates
INFO:     graphchi_engine.hpp(run:899): Finished updates
INFO:     graphchi_engine.hpp(run:852): 69.1963s: Starting: 1030932 -- 1192969
INFO:     graphchi_engine.hpp(run:865): Iteration 0/9, subinterval: 1030932 - 1192969
DEBUG:    graphchi_engine.hpp(run:880): Allocation 162038 vertices, sizeof:64 total:10370432
DEBUG:    memoryshard.hpp(load_edata:327): Compressed/full size: 0.036625 number of blocks: 93
ERROR:    stripedio.hpp(open_session:404): Could not open: rrall.edata..Z.e8B.13_15_blockdir_1048576/91 session: 2713 error: Too many open files
communitydetection: ./src/io/stripedio.hpp:406: int graphchi::stripedio::open_session(std::string, bool, bool): Assertion `rddesc>=0' failed.

Very appreciated if you could give me some hints.

Xiao

--
You received this message because you are subscribed to the Google Groups "GraphLab Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to graphlab-kdd...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.