1 .TH HBAL 1 2009-03-23 htools "Ganeti H-tools"
3 hbal \- Cluster balancer for Ganeti
12 .BI "[-m " cluster "]"
13 .BI "[-n " nodes-file " ]"
14 .BI "[-i " instances-file "]"
20 hbal is a cluster balancer that looks at the current state of the
21 cluster (nodes with their total and free disk, memory, etc.) and
22 instance placement and computes a series of steps designed to bring
23 the cluster into a better state.
25 The algorithm to do so is designed to be stable (i.e. it will give you
26 the same results when restarting it from the middle of the solution)
27 and reasonably fast. It is not, however, designed to be a perfect
28 algorithm - it is possible to make it go into a corner from which it
29 can find no improvement, because it only look one "step" ahead.
31 By default, the program will show the solution incrementally as it is
32 computed, in a somewhat cryptic format; for getting the actual Ganeti
33 command list, use the \fB-C\fR option.
37 The program works in independent steps; at each step, we compute the
38 best instance move that lowers the cluster score.
40 The possible move type for an instance are combinations of
41 failover/migrate and replace-disks such that we change one of the
42 instance nodes, and the other one remains (but possibly with changed
43 role, e.g. from primary it becomes secondary). The list is:
53 replace primary, a composite move (f, r, f)
56 failover and replace secondary, also composite (f, r)
59 replace secondary and failover, also composite (r, f)
62 We don't do the only remaining possibility of replacing both nodes
63 (r,f,r,f or the equivalent f,r,f,r) since these move needs an
64 exhaustive search over both candidate primary and secondary nodes, and
65 is O(n*n) in the number of nodes. Furthermore, it doesn't seems to
66 give better scores but will result in more disk replacements.
70 As said before, the algorithm tries to minimise the cluster score at
71 each step. Currently this score is computed as a sum of the following
76 coefficient of variance of the percent of free memory
79 coefficient of variance of the percent of reserved memory
82 coefficient of variance of the percent of free disk
85 percentage of nodes failing N+1 check
88 percentage of instances living (either as primary or secondary) on
92 The free memory and free disk values help ensure that all nodes are
93 somewhat balanced in their resource usage. The reserved memory helps
94 to ensure that nodes are somewhat balanced in holding secondary
95 instances, and that no node keeps too much memory reserved for
96 N+1. And finally, the N+1 percentage helps guide the algorithm towards
97 eliminating N+1 failures, if possible.
99 Except for the N+1 failures and offline instances percentage, we use
100 the coefficient of variance since this brings the values into the same
101 unit so to speak, and with a restrict domain of values (between zero
102 and one). The percentage of N+1 failures, while also in this numeric
103 range, doesn't actually has the same meaning, but it has shown to work
106 The other alternative, using for N+1 checks the coefficient of
107 variance of (N+1 fail=1, N+1 pass=0) across nodes could hint the
108 algorithm to make more N+1 failures if most nodes are N+1 fail
109 already. Since this (making N+1 failures) is not allowed by other
110 rules of the algorithm, so the N+1 checks would simply not work
111 anymore in this case.
113 The offline instances percentage (meaning the percentage of instances
114 living on offline nodes) will cause the algorithm to actively move
115 instances away from offline nodes. This, coupled with the restriction
116 on placement given by offline nodes, will cause evacuation of such
119 On a perfectly balanced cluster (all nodes the same size, all
120 instances the same size and spread across the nodes equally), all
121 values would be zero. This doesn't happen too often in practice :)
123 .SS OFFLINE INSTANCES
125 Since current Ganeti versions do not report the memory used by offline
126 (down) instances, ignoring the run status of instances will cause
127 wrong calculations. For this reason, the algorithm subtracts the
128 memory size of down instances from the free node memory of their
129 primary node, in effect simulating the startup of such instances.
131 .SS OTHER POSSIBLE METRICS
133 It would be desirable to add more metrics to the algorithm, especially
134 dynamically-computed metrics, such as:
138 CPU usage of instances, combined with VCPU versus PCPU count
148 The options that can be passed to the program are as follows:
150 .B -C, --print-commands
151 Print the command list at the end of the run. Without this, the
152 program will only show a shorter, but cryptic output.
155 Prints the before and after node status, in a format designed to allow
156 the user to understand the node's most important parameters.
158 The node list will contain these informations:
162 a character denoting the status of the node, with '-' meaning an
163 offline node, '*' meaning N+1 failure and blank meaning a good node
169 the total node memory
172 the memory used by the node itself
175 the memory used by instances
178 amount memory which seems to be in use but cannot be determined why or
179 by which instance; usually this means that the hypervisor has some
180 overhead or that there are other reporting errors
186 the reserved node memory, which is the amount of free memory needed
196 number of primary instances
199 number of secondary instances
202 percent of free memory
210 Only shows a one-line output from the program, designed for the case
211 when one wants to look at multiple clusters at once and check their
214 The line will contain four fields:
219 initial cluster score
222 number of steps in the solution
228 improvement in the cluster score
234 This option (which can be given multiple times) will mark nodes as
235 being \fIoffline\fR. This means a couple of things:
240 instances won't be placed on these nodes, not even temporarily;
241 e.g. the \fIreplace primary\fR move is not available if the secondary
242 node is offline, since this move requires a failover.
245 these nodes will not be included in the score calculation (except for
246 the percentage of instances on offline nodes)
251 .BI "-n" nodefile ", --nodes=" nodefile
252 The name of the file holding node information (if not collecting via
253 RAPI), instead of the default \fInodes\fR file (but see below how to
254 customize the default value via the environment).
257 .BI "-i" instancefile ", --instances=" instancefile
258 The name of the file holding instance information (if not collecting
259 via RAPI), instead of the default \fIinstances\fR file (but see below
260 how to customize the default value via the environment).
264 Collect data not from files but directly from the
266 given as an argument via RAPI. This work for both Ganeti 1.2 and
270 .BI "-l" N ", --max-length=" N
271 Restrict the solution to this length. This can be used for example to
272 automate the execution of the balancing.
276 Increase the output verbosity. Each usage of this option will increase
277 the verbosity (currently more than 2 doesn't make sense) from the
282 Just show the program version and exit.
286 The exist status of the command will be zero, unless for some reason
287 the algorithm fatally failed (e.g. wrong node or instance data).
291 If the variables \fBHTOOLS_NODES\fR and \fBHTOOLS_INSTANCES\fR are
292 present in the environment, they will override the default names for
293 the nodes and instances files. These will have of course no effect
298 The program does not check its input data for consistency, and aborts
299 with cryptic errors messages in this case.
301 The algorithm is not perfect.
303 The algorithm doesn't deal with non-\fBdrbd\fR instances, and chokes
304 on input data which has such instances.
306 The output format is not easily scriptable, and the program should
307 feed moves directly into Ganeti (either via RAPI or via a gnt-debug
312 Note that this example are not for the latest version (they don't have
317 With the default options, the program shows each individual step and
318 the improvements it brings in cluster score:
323 Loaded 20 nodes, 80 instances
324 Cluster is not N+1 happy, continuing but no guarantee that the cluster will end N+1 happy.
325 Initial score: 0.52329131
326 Trying to minimize the CV...
327 1. instance14 node1:node10 => node16:node10 0.42109120 a=f r:node16 f
328 2. instance54 node4:node15 => node16:node15 0.31904594 a=f r:node16 f
329 3. instance4 node5:node2 => node2:node16 0.26611015 a=f r:node16
330 4. instance48 node18:node20 => node2:node18 0.21361717 a=r:node2 f
331 5. instance93 node19:node18 => node16:node19 0.16166425 a=r:node16 f
332 6. instance89 node3:node20 => node2:node3 0.11005629 a=r:node2 f
333 7. instance5 node6:node2 => node16:node6 0.05841589 a=r:node16 f
334 8. instance94 node7:node20 => node20:node16 0.00658759 a=f r:node16
335 9. instance44 node20:node2 => node2:node15 0.00438740 a=f r:node15
336 10. instance62 node14:node18 => node14:node16 0.00390087 a=r:node16
337 11. instance13 node11:node14 => node11:node16 0.00361787 a=r:node16
338 12. instance19 node10:node11 => node10:node7 0.00336636 a=r:node7
339 13. instance43 node12:node13 => node12:node1 0.00305681 a=r:node1
340 14. instance1 node1:node2 => node1:node4 0.00263124 a=r:node4
341 15. instance58 node19:node20 => node19:node17 0.00252594 a=r:node17
342 Cluster score improved from 0.52329131 to 0.00252594
346 In the above output, we can see:
347 - the input data (here from files) shows a cluster with 20 nodes and
349 - the cluster is not initially N+1 compliant
350 - the initial score is 0.52329131
352 The step list follows, showing the instance, its initial
353 primary/secondary nodes, the new primary secondary, the cluster list,
354 and the actions taken in this step (with 'f' denoting failover/migrate
355 and 'r' denoting replace secondary).
357 Finally, the program shows the improvement in cluster score.
359 A more detailed output is obtained via the \fB-C\fR and \fB-p\fR options:
364 Loaded 20 nodes, 80 instances
365 Cluster is not N+1 happy, continuing but no guarantee that the cluster will end N+1 happy.
366 Initial cluster status:
367 N1 Name t_mem f_mem r_mem t_dsk f_dsk pri sec p_fmem p_fdsk
368 * node1 32762 1280 6000 1861 1026 5 3 0.03907 0.55179
369 node2 32762 31280 12000 1861 1026 0 8 0.95476 0.55179
370 * node3 32762 1280 6000 1861 1026 5 3 0.03907 0.55179
371 * node4 32762 1280 6000 1861 1026 5 3 0.03907 0.55179
372 * node5 32762 1280 6000 1861 978 5 5 0.03907 0.52573
373 * node6 32762 1280 6000 1861 1026 5 3 0.03907 0.55179
374 * node7 32762 1280 6000 1861 1026 5 3 0.03907 0.55179
375 node8 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
376 node9 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
377 * node10 32762 7280 12000 1861 1026 4 4 0.22221 0.55179
378 node11 32762 7280 6000 1861 922 4 5 0.22221 0.49577
379 node12 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
380 node13 32762 7280 6000 1861 922 4 5 0.22221 0.49577
381 node14 32762 7280 6000 1861 922 4 5 0.22221 0.49577
382 * node15 32762 7280 12000 1861 1131 4 3 0.22221 0.60782
383 node16 32762 31280 0 1861 1860 0 0 0.95476 1.00000
384 node17 32762 7280 6000 1861 1106 5 3 0.22221 0.59479
385 * node18 32762 1280 6000 1396 561 5 3 0.03907 0.40239
386 * node19 32762 1280 6000 1861 1026 5 3 0.03907 0.55179
387 node20 32762 13280 12000 1861 689 3 9 0.40535 0.37068
389 Initial score: 0.52329131
390 Trying to minimize the CV...
391 1. instance14 node1:node10 => node16:node10 0.42109120 a=f r:node16 f
392 2. instance54 node4:node15 => node16:node15 0.31904594 a=f r:node16 f
393 3. instance4 node5:node2 => node2:node16 0.26611015 a=f r:node16
394 4. instance48 node18:node20 => node2:node18 0.21361717 a=r:node2 f
395 5. instance93 node19:node18 => node16:node19 0.16166425 a=r:node16 f
396 6. instance89 node3:node20 => node2:node3 0.11005629 a=r:node2 f
397 7. instance5 node6:node2 => node16:node6 0.05841589 a=r:node16 f
398 8. instance94 node7:node20 => node20:node16 0.00658759 a=f r:node16
399 9. instance44 node20:node2 => node2:node15 0.00438740 a=f r:node15
400 10. instance62 node14:node18 => node14:node16 0.00390087 a=r:node16
401 11. instance13 node11:node14 => node11:node16 0.00361787 a=r:node16
402 12. instance19 node10:node11 => node10:node7 0.00336636 a=r:node7
403 13. instance43 node12:node13 => node12:node1 0.00305681 a=r:node1
404 14. instance1 node1:node2 => node1:node4 0.00263124 a=r:node4
405 15. instance58 node19:node20 => node19:node17 0.00252594 a=r:node17
406 Cluster score improved from 0.52329131 to 0.00252594
408 Commands to run to reach the above solution:
410 echo gnt-instance migrate instance14
411 echo gnt-instance replace-disks -n node16 instance14
412 echo gnt-instance migrate instance14
414 echo gnt-instance migrate instance54
415 echo gnt-instance replace-disks -n node16 instance54
416 echo gnt-instance migrate instance54
418 echo gnt-instance migrate instance4
419 echo gnt-instance replace-disks -n node16 instance4
421 echo gnt-instance replace-disks -n node2 instance48
422 echo gnt-instance migrate instance48
424 echo gnt-instance replace-disks -n node16 instance93
425 echo gnt-instance migrate instance93
427 echo gnt-instance replace-disks -n node2 instance89
428 echo gnt-instance migrate instance89
430 echo gnt-instance replace-disks -n node16 instance5
431 echo gnt-instance migrate instance5
433 echo gnt-instance migrate instance94
434 echo gnt-instance replace-disks -n node16 instance94
436 echo gnt-instance migrate instance44
437 echo gnt-instance replace-disks -n node15 instance44
439 echo gnt-instance replace-disks -n node16 instance62
441 echo gnt-instance replace-disks -n node16 instance13
443 echo gnt-instance replace-disks -n node7 instance19
445 echo gnt-instance replace-disks -n node1 instance43
447 echo gnt-instance replace-disks -n node4 instance1
449 echo gnt-instance replace-disks -n node17 instance58
451 Final cluster status:
452 N1 Name t_mem f_mem r_mem t_dsk f_dsk pri sec p_fmem p_fdsk
453 node1 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
454 node2 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
455 node3 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
456 node4 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
457 node5 32762 7280 6000 1861 1078 4 5 0.22221 0.57947
458 node6 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
459 node7 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
460 node8 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
461 node9 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
462 node10 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
463 node11 32762 7280 6000 1861 1022 4 4 0.22221 0.54951
464 node12 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
465 node13 32762 7280 6000 1861 1022 4 4 0.22221 0.54951
466 node14 32762 7280 6000 1861 1022 4 4 0.22221 0.54951
467 node15 32762 7280 6000 1861 1031 4 4 0.22221 0.55408
468 node16 32762 7280 6000 1861 1060 4 4 0.22221 0.57007
469 node17 32762 7280 6000 1861 1006 5 4 0.22221 0.54105
470 node18 32762 7280 6000 1396 761 4 2 0.22221 0.54570
471 node19 32762 7280 6000 1861 1026 4 4 0.22221 0.55179
472 node20 32762 13280 6000 1861 1089 3 5 0.40535 0.58565
477 Here we see, beside the step list, the initial and final cluster
478 status, with the final one showing all nodes being N+1 compliant, and
479 the command list to reach the final solution. In the initial listing,
480 we see which nodes are not N+1 compliant.
482 The algorithm is stable as long as each step above is fully completed,
483 e.g. in step 8, both the migrate and the replace-disks are
484 done. Otherwise, if only the migrate is done, the input data is
485 changed in a way that the program will output a different solution
486 list (but hopefully will end in the same state).
489 .BR hn1 "(1), " hscan "(1), " ganeti "(7), " gnt-instance "(8), "