]>
Commit | Line | Data |
---|---|---|
4177a14d TL |
1 | package PVE::CLI::pve6to7; |
2 | ||
3 | use strict; | |
4 | use warnings; | |
5 | ||
6 | use PVE::API2::APT; | |
7 | use PVE::API2::Ceph; | |
8 | use PVE::API2::LXC; | |
9 | use PVE::API2::Qemu; | |
10 | use PVE::API2::Certificates; | |
11 | ||
12 | use PVE::Ceph::Tools; | |
13 | use PVE::Cluster; | |
14 | use PVE::Corosync; | |
15 | use PVE::INotify; | |
16 | use PVE::JSONSchema; | |
17 | use PVE::RPCEnvironment; | |
18 | use PVE::Storage; | |
19 | use PVE::Tools qw(run_command); | |
20 | use PVE::QemuServer; | |
e522a2db | 21 | use PVE::VZDump::Common; |
4177a14d TL |
22 | |
23 | use Term::ANSIColor; | |
24 | ||
25 | use PVE::CLIHandler; | |
26 | ||
27 | use base qw(PVE::CLIHandler); | |
28 | ||
29 | my $nodename = PVE::INotify::nodename(); | |
30 | ||
31 | sub setup_environment { | |
32 | PVE::RPCEnvironment->setup_default_cli_env(); | |
33 | } | |
34 | ||
35 | my $min_pve_major = 6; | |
36 | my $min_pve_minor = 4; | |
37 | my $min_pve_pkgrel = 1; | |
38 | ||
39 | my $counters = { | |
40 | pass => 0, | |
41 | skip => 0, | |
42 | warn => 0, | |
43 | fail => 0, | |
44 | }; | |
45 | ||
46 | my $log_line = sub { | |
47 | my ($level, $line) = @_; | |
48 | ||
49 | $counters->{$level}++ if defined($level) && defined($counters->{$level}); | |
50 | ||
51 | print uc($level), ': ' if defined($level); | |
52 | print "$line\n"; | |
53 | }; | |
54 | ||
55 | sub log_pass { | |
56 | print color('green'); | |
57 | $log_line->('pass', @_); | |
58 | print color('reset'); | |
59 | } | |
60 | ||
61 | sub log_info { | |
62 | $log_line->('info', @_); | |
63 | } | |
64 | sub log_skip { | |
65 | $log_line->('skip', @_); | |
66 | } | |
67 | sub log_warn { | |
68 | print color('yellow'); | |
69 | $log_line->('warn', @_); | |
70 | print color('reset'); | |
71 | } | |
72 | sub log_fail { | |
73 | print color('red'); | |
74 | $log_line->('fail', @_); | |
75 | print color('reset'); | |
76 | } | |
77 | ||
78 | my $print_header_first = 1; | |
79 | sub print_header { | |
80 | my ($h) = @_; | |
81 | print "\n" if !$print_header_first; | |
82 | print "= $h =\n\n"; | |
83 | $print_header_first = 0; | |
84 | } | |
85 | ||
86 | my $get_systemd_unit_state = sub { | |
87 | my ($unit) = @_; | |
88 | ||
89 | my $state; | |
90 | my $filter_output = sub { | |
91 | $state = shift; | |
92 | chomp $state; | |
93 | }; | |
94 | eval { | |
95 | run_command(['systemctl', 'is-enabled', "$unit"], outfunc => $filter_output, noerr => 1); | |
96 | return if !defined($state); | |
97 | run_command(['systemctl', 'is-active', "$unit"], outfunc => $filter_output, noerr => 1); | |
98 | }; | |
99 | ||
100 | return $state // 'unknown'; | |
101 | }; | |
102 | my $log_systemd_unit_state = sub { | |
103 | my ($unit, $no_fail_on_inactive) = @_; | |
104 | ||
105 | my $log_method = \&log_warn; | |
106 | ||
107 | my $state = $get_systemd_unit_state->($unit); | |
108 | if ($state eq 'active') { | |
109 | $log_method = \&log_pass; | |
110 | } elsif ($state eq 'inactive') { | |
111 | $log_method = $no_fail_on_inactive ? \&log_warn : \&log_fail; | |
112 | } elsif ($state eq 'failed') { | |
113 | $log_method = \&log_fail; | |
114 | } | |
115 | ||
116 | $log_method->("systemd unit '$unit' is in state '$state'"); | |
117 | }; | |
118 | ||
119 | my $versions; | |
120 | my $get_pkg = sub { | |
121 | my ($pkg) = @_; | |
122 | ||
123 | $versions = eval { PVE::API2::APT->versions({ node => $nodename }) } if !defined($versions); | |
124 | ||
125 | if (!defined($versions)) { | |
126 | my $msg = "unable to retrieve package version information"; | |
127 | $msg .= "- $@" if $@; | |
128 | log_fail("$msg"); | |
129 | return undef; | |
130 | } | |
131 | ||
132 | my $pkgs = [ grep { $_->{Package} eq $pkg } @$versions ]; | |
133 | if (!defined $pkgs || $pkgs == 0) { | |
134 | log_fail("unable to determine installed $pkg version."); | |
135 | return undef; | |
136 | } else { | |
137 | return $pkgs->[0]; | |
138 | } | |
139 | }; | |
140 | ||
141 | sub check_pve_packages { | |
142 | print_header("CHECKING VERSION INFORMATION FOR PVE PACKAGES"); | |
143 | ||
144 | print "Checking for package updates..\n"; | |
145 | my $updates = eval { PVE::API2::APT->list_updates({ node => $nodename }); }; | |
146 | if (!defined($updates)) { | |
147 | log_warn("$@") if $@; | |
148 | log_fail("unable to retrieve list of package updates!"); | |
149 | } elsif (@$updates > 0) { | |
150 | my $pkgs = join(', ', map { $_->{Package} } @$updates); | |
151 | log_warn("updates for the following packages are available:\n $pkgs"); | |
152 | } else { | |
153 | log_pass("all packages uptodate"); | |
154 | } | |
155 | ||
156 | print "\nChecking proxmox-ve package version..\n"; | |
157 | if (defined(my $proxmox_ve = $get_pkg->('proxmox-ve'))) { | |
158 | my $min_pve_ver = "$min_pve_major.$min_pve_minor-$min_pve_pkgrel"; | |
159 | ||
160 | my ($maj, $min, $pkgrel) = $proxmox_ve->{OldVersion} =~ m/^(\d+)\.(\d+)-(\d+)/; | |
161 | ||
162 | my $upgraded = 0; | |
163 | ||
164 | if ($maj > $min_pve_major) { | |
165 | log_pass("already upgraded to Proxmox VE " . ($min_pve_major + 1)); | |
166 | $upgraded = 1; | |
167 | } elsif ($maj >= $min_pve_major && $min >= $min_pve_minor && $pkgrel >= $min_pve_pkgrel) { | |
168 | log_pass("proxmox-ve package has version >= $min_pve_ver"); | |
169 | } else { | |
170 | log_fail("proxmox-ve package is too old, please upgrade to >= $min_pve_ver!"); | |
171 | } | |
172 | ||
173 | my ($krunning, $kinstalled) = (qr/5\.11/, 'pve-kernel-5.11'); | |
174 | if (!$upgraded) { | |
175 | ($krunning, $kinstalled) = (qr/5\.(?:4|11)/, 'pve-kernel-4.15'); | |
176 | } | |
177 | ||
178 | print "\nChecking running kernel version..\n"; | |
179 | my $kernel_ver = $proxmox_ve->{RunningKernel}; | |
180 | if (!defined($kernel_ver)) { | |
181 | log_fail("unable to determine running kernel version."); | |
182 | } elsif ($kernel_ver =~ /^$krunning/) { | |
183 | log_pass("expected running kernel '$kernel_ver'."); | |
184 | } elsif ($get_pkg->($kinstalled)) { | |
185 | log_warn("expected kernel '$kinstalled' intalled but not yet rebooted!"); | |
186 | } else { | |
187 | log_warn("unexpected running and installed kernel '$kernel_ver'."); | |
188 | } | |
189 | } else { | |
190 | log_fail("proxmox-ve package not found!"); | |
191 | } | |
192 | } | |
193 | ||
194 | ||
195 | sub check_storage_health { | |
196 | print_header("CHECKING CONFIGURED STORAGES"); | |
197 | my $cfg = PVE::Storage::config(); | |
198 | ||
199 | my $ctime = time(); | |
200 | ||
201 | my $info = PVE::Storage::storage_info($cfg); | |
202 | ||
203 | foreach my $storeid (keys %$info) { | |
204 | my $d = $info->{$storeid}; | |
205 | if ($d->{enabled}) { | |
206 | if ($d->{type} eq 'sheepdog') { | |
207 | log_fail("storage '$storeid' of type 'sheepdog' is enabled - experimental sheepdog support dropped in PVE 6") | |
208 | } elsif ($d->{active}) { | |
209 | log_pass("storage '$storeid' enabled and active."); | |
210 | } else { | |
211 | log_warn("storage '$storeid' enabled but not active!"); | |
212 | } | |
213 | } else { | |
214 | log_skip("storage '$storeid' disabled."); | |
215 | } | |
216 | } | |
217 | } | |
218 | ||
219 | sub check_cluster_corosync { | |
220 | print_header("CHECKING CLUSTER HEALTH/SETTINGS"); | |
221 | ||
222 | if (!PVE::Corosync::check_conf_exists(1)) { | |
223 | log_skip("standalone node."); | |
224 | return; | |
225 | } | |
226 | ||
227 | $log_systemd_unit_state->('pve-cluster.service'); | |
228 | $log_systemd_unit_state->('corosync.service'); | |
229 | ||
230 | if (PVE::Cluster::check_cfs_quorum(1)) { | |
231 | log_pass("Cluster Filesystem is quorate."); | |
232 | } else { | |
233 | log_fail("Cluster Filesystem readonly, lost quorum?!"); | |
234 | } | |
235 | ||
236 | my $conf = PVE::Cluster::cfs_read_file('corosync.conf'); | |
237 | my $conf_nodelist = PVE::Corosync::nodelist($conf); | |
238 | my $node_votes = 0; | |
239 | ||
240 | print "\nAnalzying quorum settings and state..\n"; | |
241 | if (!defined($conf_nodelist)) { | |
242 | log_fail("unable to retrieve nodelist from corosync.conf"); | |
243 | } else { | |
244 | if (grep { $conf_nodelist->{$_}->{quorum_votes} != 1 } keys %$conf_nodelist) { | |
245 | log_warn("non-default quorum_votes distribution detected!"); | |
246 | } | |
247 | map { $node_votes += $conf_nodelist->{$_}->{quorum_votes} // 0 } keys %$conf_nodelist; | |
248 | } | |
249 | ||
250 | my ($expected_votes, $total_votes); | |
251 | my $filter_output = sub { | |
252 | my $line = shift; | |
253 | ($expected_votes) = $line =~ /^Expected votes:\s*(\d+)\s*$/ | |
254 | if !defined($expected_votes); | |
255 | ($total_votes) = $line =~ /^Total votes:\s*(\d+)\s*$/ | |
256 | if !defined($total_votes); | |
257 | }; | |
258 | eval { | |
259 | run_command(['corosync-quorumtool', '-s'], outfunc => $filter_output, noerr => 1); | |
260 | }; | |
261 | ||
262 | if (!defined($expected_votes)) { | |
263 | log_fail("unable to get expected number of votes, setting to 0."); | |
264 | $expected_votes = 0; | |
265 | } | |
266 | if (!defined($total_votes)) { | |
267 | log_fail("unable to get expected number of votes, setting to 0."); | |
268 | $total_votes = 0; | |
269 | } | |
270 | ||
271 | my $cfs_nodelist = PVE::Cluster::get_clinfo()->{nodelist}; | |
272 | my $offline_nodes = grep { $cfs_nodelist->{$_}->{online} != 1 } keys %$cfs_nodelist; | |
273 | if ($offline_nodes > 0) { | |
274 | log_fail("$offline_nodes nodes are offline!"); | |
275 | } | |
276 | ||
277 | my $qdevice_votes = 0; | |
278 | if (my $qdevice_setup = $conf->{main}->{quorum}->{device}) { | |
279 | $qdevice_votes = $qdevice_setup->{votes} // 1; | |
280 | } | |
281 | ||
282 | log_info("configured votes - nodes: $node_votes"); | |
283 | log_info("configured votes - qdevice: $qdevice_votes"); | |
284 | log_info("current expected votes: $expected_votes"); | |
285 | log_info("current total votes: $total_votes"); | |
286 | ||
287 | log_warn("expected votes set to non-standard value '$expected_votes'.") | |
288 | if $expected_votes != $node_votes + $qdevice_votes; | |
289 | log_warn("total votes < expected votes: $total_votes/$expected_votes!") | |
290 | if $total_votes < $expected_votes; | |
291 | ||
292 | my $conf_nodelist_count = scalar(keys %$conf_nodelist); | |
293 | my $cfs_nodelist_count = scalar(keys %$cfs_nodelist); | |
294 | log_warn("cluster consists of less than three quorum-providing nodes!") | |
295 | if $conf_nodelist_count < 3 && $conf_nodelist_count + $qdevice_votes < 3; | |
296 | ||
297 | log_fail("corosync.conf ($conf_nodelist_count) and pmxcfs ($cfs_nodelist_count) don't agree about size of nodelist.") | |
298 | if $conf_nodelist_count != $cfs_nodelist_count; | |
299 | ||
300 | print "\nChecking nodelist entries..\n"; | |
301 | for my $cs_node (sort keys %$conf_nodelist) { | |
302 | my $entry = $conf_nodelist->{$cs_node}; | |
303 | log_fail("$cs_node: no name entry in corosync.conf.") | |
304 | if !defined($entry->{name}); | |
305 | log_fail("$cs_node: no nodeid configured in corosync.conf.") | |
306 | if !defined($entry->{nodeid}); | |
307 | my $gotLinks = 0; | |
308 | for my $link (0..7) { | |
309 | $gotLinks++ if defined($entry->{"ring${link}_addr"}); | |
310 | } | |
311 | log_fail("$cs_node: no ringX_addr (0 <= X <= 7) link defined in corosync.conf.") if $gotLinks <= 0; | |
312 | ||
313 | my $verify_ring_ip = sub { | |
314 | my $key = shift; | |
315 | if (defined(my $ring = $entry->{$key})) { | |
316 | my ($resolved_ip, undef) = PVE::Corosync::resolve_hostname_like_corosync($ring, $conf); | |
317 | if (defined($resolved_ip)) { | |
318 | if ($resolved_ip ne $ring) { | |
319 | log_warn("$cs_node: $key '$ring' resolves to '$resolved_ip'.\n Consider replacing it with the currently resolved IP address."); | |
320 | } else { | |
321 | log_pass("$cs_node: $key is configured to use IP address '$ring'"); | |
322 | } | |
323 | } else { | |
324 | log_fail("$cs_node: unable to resolve $key '$ring' to an IP address according to Corosync's resolve strategy - cluster will potentially fail with Corosync 3.x/kronosnet!"); | |
325 | } | |
326 | } | |
327 | }; | |
328 | for my $link (0..7) { | |
329 | $verify_ring_ip->("ring${link}_addr"); | |
330 | } | |
331 | } | |
332 | ||
333 | print "\nChecking totem settings..\n"; | |
334 | my $totem = $conf->{main}->{totem}; | |
335 | my $transport = $totem->{transport}; | |
336 | if (defined($transport)) { | |
337 | if ($transport ne 'knet') { | |
338 | log_fail("Corosync transport explicitly set to '$transport' instead of implicit default!"); | |
339 | } else { | |
340 | log_pass("Corosync transport set to '$transport'."); | |
341 | } | |
342 | } else { | |
343 | log_pass("Corosync transport set to implicit default."); | |
344 | } | |
345 | ||
346 | # TODO: are those values still up-to-date? | |
347 | if ((!defined($totem->{secauth}) || $totem->{secauth} ne 'on') && (!defined($totem->{crypto_cipher}) || $totem->{crypto_cipher} eq 'none')) { | |
348 | log_fail("Corosync authentication/encryption is not explicitly enabled (secauth / crypto_cipher / crypto_hash)!"); | |
349 | } else { | |
350 | if (defined($totem->{crypto_cipher}) && $totem->{crypto_cipher} eq '3des') { | |
351 | log_fail("Corosync encryption cipher set to '3des', no longer supported in Corosync 3.x!"); # FIXME: can be removed? | |
352 | } else { | |
353 | log_pass("Corosync encryption and authentication enabled."); | |
354 | } | |
355 | } | |
356 | ||
357 | print "\n"; | |
358 | log_info("run 'pvecm status' to get detailed cluster status.."); | |
359 | ||
360 | print_header("CHECKING INSTALLED COROSYNC VERSION"); | |
361 | if (defined(my $corosync = $get_pkg->('corosync'))) { | |
362 | if ($corosync->{OldVersion} =~ m/^2\./) { | |
363 | log_fail("corosync 2.x installed, cluster-wide upgrade to 3.x needed!"); | |
364 | } elsif ($corosync->{OldVersion} =~ m/^3\./) { | |
365 | log_pass("corosync 3.x installed."); | |
366 | } else { | |
367 | log_fail("unexpected corosync version installed: $corosync->{OldVersion}!"); | |
368 | } | |
369 | } | |
370 | } | |
371 | ||
372 | sub check_ceph { | |
373 | print_header("CHECKING HYPER-CONVERGED CEPH STATUS"); | |
374 | ||
375 | if (PVE::Ceph::Tools::check_ceph_inited(1)) { | |
376 | log_info("hyper-converged ceph setup detected!"); | |
377 | } else { | |
378 | log_skip("no hyper-converged ceph setup detected!"); | |
379 | return; | |
380 | } | |
381 | ||
382 | log_info("getting Ceph status/health information.."); | |
383 | my $ceph_status = eval { PVE::API2::Ceph->status({ node => $nodename }); }; | |
384 | my $osd_flags = eval { PVE::API2::Ceph->get_flags({ node => $nodename }); }; | |
385 | my $noout_wanted = 1; | |
386 | my $noout = $osd_flags && $osd_flags =~ m/noout/; | |
387 | ||
388 | if (!$ceph_status || !$ceph_status->{health}) { | |
389 | log_fail("unable to determine Ceph status!"); | |
390 | } else { | |
391 | my $ceph_health = $ceph_status->{health}->{status}; | |
392 | if (!$ceph_health) { | |
393 | log_fail("unable to determine Ceph health!"); | |
394 | } elsif ($ceph_health eq 'HEALTH_OK') { | |
395 | log_pass("Ceph health reported as 'HEALTH_OK'."); | |
396 | } elsif ($ceph_health eq 'HEALTH_WARN' && $noout && (keys %{$ceph_status->{health}->{checks}} == 1)) { | |
397 | log_pass("Ceph health reported as 'HEALTH_WARN' with a single failing check and 'noout' flag set."); | |
398 | } else { | |
399 | log_warn("Ceph health reported as '$ceph_health'.\n Use the PVE ". | |
400 | "dashboard or 'ceph -s' to determine the specific issues and try to resolve them."); | |
401 | } | |
402 | } | |
403 | ||
404 | log_info("getting Ceph OSD flags.."); | |
405 | eval { | |
406 | if (!$osd_flags) { | |
407 | log_fail("unable to get Ceph OSD flags!"); | |
408 | } else { | |
409 | if ($osd_flags =~ m/recovery_deletes/ && $osd_flags =~ m/purged_snapdirs/) { | |
410 | log_pass("all PGs have been scrubbed at least once while running Ceph Luminous."); # FIXME: remove? | |
411 | } else { | |
412 | log_fail("missing 'recovery_deletes' and/or 'purged_snapdirs' flag, scrub of all PGs required before upgrading to Nautilus!"); | |
413 | } | |
414 | } | |
415 | }; | |
416 | ||
417 | # TODO: check OSD min-required version, if to low it breaks stuff! | |
418 | ||
419 | log_info("getting Ceph daemon versions.."); | |
420 | my $ceph_versions = eval { PVE::Ceph::Tools::get_cluster_versions(undef, 1); }; | |
421 | if (!$ceph_versions) { | |
422 | log_fail("unable to determine Ceph daemon versions!"); | |
423 | } else { | |
424 | my $services = [ | |
425 | { 'key' => 'mon', 'name' => 'monitor' }, | |
426 | { 'key' => 'mgr', 'name' => 'manager' }, | |
427 | { 'key' => 'mds', 'name' => 'MDS' }, | |
428 | { 'key' => 'osd', 'name' => 'OSD' }, | |
429 | ]; | |
430 | ||
431 | foreach my $service (@$services) { | |
432 | my $name = $service->{name}; | |
433 | if (my $service_versions = $ceph_versions->{$service->{key}}) { | |
434 | if (keys %$service_versions == 0) { | |
435 | log_skip("no running instances detected for daemon type $name."); | |
436 | } elsif (keys %$service_versions == 1) { | |
437 | log_pass("single running version detected for daemon type $name."); | |
438 | } else { | |
439 | log_warn("multiple running versions detected for daemon type $name!"); | |
440 | } | |
441 | } else { | |
442 | log_skip("unable to determine versions of running Ceph $name instances."); | |
443 | } | |
444 | } | |
445 | ||
446 | my $overall_versions = $ceph_versions->{overall}; | |
447 | if (!$overall_versions) { | |
448 | log_warn("unable to determine overall Ceph daemon versions!"); | |
449 | } elsif (keys %$overall_versions == 1) { | |
450 | log_pass("single running overall version detected for all Ceph daemon types."); | |
451 | if ((keys %$overall_versions)[0] =~ /^ceph version 15\./) { | |
452 | $noout_wanted = 0; | |
453 | } | |
454 | } else { | |
455 | log_warn("overall version mismatch detected, check 'ceph versions' output for details!"); | |
456 | } | |
457 | } | |
458 | ||
459 | if ($noout) { | |
460 | if ($noout_wanted) { | |
461 | log_pass("'noout' flag set to prevent rebalancing during cluster-wide upgrades."); | |
462 | } else { | |
463 | log_warn("'noout' flag set, Ceph cluster upgrade seems finished."); | |
464 | } | |
465 | } elsif ($noout_wanted) { | |
466 | log_warn("'noout' flag not set - recommended to prevent rebalancing during upgrades."); | |
467 | } | |
468 | ||
469 | log_info("checking Ceph config.."); | |
470 | my $conf = PVE::Cluster::cfs_read_file('ceph.conf'); | |
471 | if (%$conf) { | |
472 | my $global = $conf->{global}; | |
473 | ||
474 | my $global_monhost = $global->{mon_host} // $global->{"mon host"} // $global->{"mon-host"}; | |
475 | if (!defined($global_monhost)) { | |
476 | log_warn("No 'mon_host' entry found in ceph config.\n It's recommended to add mon_host with all monitor addresses (without ports) to the global section."); | |
477 | } else { | |
478 | log_pass("Found 'mon_host' entry."); | |
479 | } | |
480 | ||
481 | my $ipv6 = $global->{ms_bind_ipv6} // $global->{"ms bind ipv6"} // $global->{"ms-bind-ipv6"}; | |
482 | if ($ipv6) { | |
483 | my $ipv4 = $global->{ms_bind_ipv4} // $global->{"ms bind ipv4"} // $global->{"ms-bind-ipv4"}; | |
484 | if ($ipv6 eq 'true' && (!defined($ipv4) || $ipv4 ne 'false')) { | |
485 | log_warn("'ms_bind_ipv6' is enabled but 'ms_bind_ipv4' is not disabled.\n Make sure to disable 'ms_bind_ipv4' for ipv6 only clusters, or add an ipv4 network to public/cluster network."); | |
486 | } else { | |
487 | log_pass("'ms_bind_ipv6' is enabled and 'ms_bind_ipv4' disabled"); | |
488 | } | |
489 | } else { | |
490 | log_pass("'ms_bind_ipv6' not enabled"); | |
491 | } | |
492 | ||
493 | if (defined($global->{keyring})) { | |
494 | log_warn("[global] config section contains 'keyring' option, which will prevent services from starting with Nautilus.\n Move 'keyring' option to [client] section instead."); | |
495 | } else { | |
496 | log_pass("no 'keyring' option in [global] section found."); | |
497 | } | |
498 | ||
499 | } else { | |
500 | log_warn("Empty ceph config found"); | |
501 | } | |
502 | ||
503 | my $local_ceph_ver = PVE::Ceph::Tools::get_local_version(1); | |
504 | if (defined($local_ceph_ver)) { | |
505 | if ($local_ceph_ver == 14) { | |
506 | my $ceph_volume_osds = PVE::Ceph::Tools::ceph_volume_list(); | |
507 | my $scanned_osds = PVE::Tools::dir_glob_regex('/etc/ceph/osd', '^.*\.json$'); | |
508 | if (-e '/var/lib/ceph/osd/' && !defined($scanned_osds) && !(keys %$ceph_volume_osds)) { | |
509 | log_warn("local Ceph version is Nautilus, local OSDs detected, but no conversion from ceph-disk to ceph-volume done (yet)."); | |
510 | } | |
511 | } | |
512 | } else { | |
513 | log_fail("unable to determine local Ceph version."); | |
514 | } | |
515 | } | |
516 | ||
e522a2db FE |
517 | sub check_backup_retention_settings { |
518 | log_info("Checking backup retention settings.."); | |
519 | ||
520 | my $pass = 1; | |
521 | ||
522 | my $node_has_retention; | |
523 | ||
524 | my $maxfiles_msg = "parameter 'maxfiles' is deprecated with PVE 7.x and will be removed in a " . | |
525 | "future version, use 'prune-backups' instead."; | |
526 | ||
527 | eval { | |
528 | my $confdesc = PVE::VZDump::Common::get_confdesc(); | |
529 | ||
530 | my $fn = "/etc/vzdump.conf"; | |
531 | my $raw = PVE::Tools::file_get_contents($fn); | |
532 | ||
533 | my $conf_schema = { type => 'object', properties => $confdesc, }; | |
534 | my $param = PVE::JSONSchema::parse_config($conf_schema, $fn, $raw); | |
535 | ||
536 | if (defined($param->{maxfiles})) { | |
537 | $pass = 0; | |
538 | log_warn("$fn - $maxfiles_msg"); | |
539 | } | |
540 | ||
541 | $node_has_retention = defined($param->{maxfiles}) || defined($param->{'prune-backups'}); | |
542 | }; | |
543 | if (my $err = $@) { | |
544 | $pass = 0; | |
545 | log_warn("unable to parse node's VZDump configuration - $err"); | |
546 | } | |
547 | ||
548 | my $storage_cfg = PVE::Storage::config(); | |
549 | ||
550 | for my $storeid (keys $storage_cfg->{ids}->%*) { | |
551 | my $scfg = $storage_cfg->{ids}->{$storeid}; | |
552 | ||
553 | if (defined($scfg->{maxfiles})) { | |
554 | $pass = 0; | |
555 | log_warn("storage '$storeid' - $maxfiles_msg"); | |
556 | } | |
557 | ||
558 | next if !$scfg->{content}->{backup}; | |
559 | next if defined($scfg->{maxfiles}) || defined($scfg->{'prune-backups'}); | |
560 | next if $node_has_retention; | |
561 | ||
562 | log_info("storage '$storeid' - no backup retention settings defined - by default, PVE " . | |
563 | "7.x will no longer keep only the last backup, but all backups"); | |
564 | } | |
565 | ||
437ade60 FG |
566 | eval { |
567 | my $vzdump_cron = PVE::Cluster::cfs_read_file('vzdump.cron'); | |
e522a2db | 568 | |
437ade60 FG |
569 | # only warn once, there might be many jobs... |
570 | if (scalar(grep { defined($_->{maxfiles}) } $vzdump_cron->{jobs}->@*)) { | |
571 | $pass = 0; | |
572 | log_warn("/etc/pve/vzdump.cron - $maxfiles_msg"); | |
573 | } | |
574 | }; | |
575 | if (my $err = $@) { | |
e522a2db | 576 | $pass = 0; |
437ade60 | 577 | log_warn("unable to parse node's VZDump configuration - $err"); |
e522a2db FE |
578 | } |
579 | ||
580 | log_pass("no problems found.") if $pass; | |
581 | } | |
582 | ||
5df8b555 FE |
583 | sub check_cifs_credential_location { |
584 | log_info("checking CIFS credential location.."); | |
585 | ||
586 | my $regex = qr/^(.*)\.cred$/; | |
587 | ||
588 | my $found; | |
589 | ||
590 | PVE::Tools::dir_glob_foreach('/etc/pve/priv/', $regex, sub { | |
591 | my ($filename) = @_; | |
592 | ||
593 | my ($basename) = $filename =~ $regex; | |
594 | ||
595 | log_warn("CIFS credentials '/etc/pve/priv/$filename' will be moved to " . | |
596 | "'/etc/pve/priv/storage/$basename.pw' during the update"); | |
597 | ||
598 | $found = 1; | |
599 | }); | |
600 | ||
601 | log_pass("no CIFS credentials at outdated location found.") if !$found; | |
602 | } | |
603 | ||
4177a14d TL |
604 | sub check_misc { |
605 | print_header("MISCELLANEOUS CHECKS"); | |
606 | my $ssh_config = eval { PVE::Tools::file_get_contents('/root/.ssh/config') }; | |
607 | if (defined($ssh_config)) { | |
608 | log_fail("Unsupported SSH Cipher configured for root in /root/.ssh/config: $1") | |
609 | if $ssh_config =~ /^Ciphers .*(blowfish|arcfour|3des).*$/m; | |
610 | } else { | |
611 | log_skip("No SSH config file found."); | |
612 | } | |
613 | ||
614 | log_info("Checking common daemon services.."); | |
615 | $log_systemd_unit_state->('pveproxy.service'); | |
616 | $log_systemd_unit_state->('pvedaemon.service'); | |
617 | $log_systemd_unit_state->('pvestatd.service'); | |
618 | ||
619 | my $root_free = PVE::Tools::df('/', 10); | |
620 | log_warn("Less than 2G free space on root file system.") | |
621 | if defined($root_free) && $root_free->{avail} < 2*1024*1024*1024; | |
622 | ||
623 | log_info("Checking for running guests.."); | |
624 | my $running_guests = 0; | |
625 | ||
626 | my $vms = eval { PVE::API2::Qemu->vmlist({ node => $nodename }) }; | |
627 | log_warn("Failed to retrieve information about this node's VMs - $@") if $@; | |
628 | $running_guests += grep { $_->{status} eq 'running' } @$vms if defined($vms); | |
629 | ||
630 | my $cts = eval { PVE::API2::LXC->vmlist({ node => $nodename }) }; | |
631 | log_warn("Failed to retrieve information about this node's CTs - $@") if $@; | |
632 | $running_guests += grep { $_->{status} eq 'running' } @$cts if defined($cts); | |
633 | ||
634 | if ($running_guests > 0) { | |
635 | log_warn("$running_guests running guest(s) detected - consider migrating or stopping them.") | |
636 | } else { | |
637 | log_pass("no running guest detected.") | |
638 | } | |
639 | ||
640 | log_info("Checking if the local node's hostname '$nodename' is resolvable.."); | |
641 | my $local_ip = eval { PVE::Network::get_ip_from_hostname($nodename) }; | |
642 | if ($@) { | |
643 | log_warn("Failed to resolve hostname '$nodename' to IP - $@"); | |
644 | } else { | |
645 | log_info("Checking if resolved IP is configured on local node.."); | |
646 | my $cidr = Net::IP::ip_is_ipv6($local_ip) ? "$local_ip/128" : "$local_ip/32"; | |
647 | my $configured_ips = PVE::Network::get_local_ip_from_cidr($cidr); | |
648 | my $ip_count = scalar(@$configured_ips); | |
649 | ||
650 | if ($ip_count <= 0) { | |
651 | log_fail("Resolved node IP '$local_ip' not configured or active for '$nodename'"); | |
652 | } elsif ($ip_count > 1) { | |
653 | log_warn("Resolved node IP '$local_ip' active on multiple ($ip_count) interfaces!"); | |
654 | } else { | |
655 | log_pass("Resolved node IP '$local_ip' configured and active on single interface."); | |
656 | } | |
657 | } | |
658 | ||
659 | log_info("Check node certificate's RSA key size"); | |
660 | my $certs = PVE::API2::Certificates->info({ node => $nodename }); | |
661 | my $certs_check = { | |
662 | 'rsaEncryption' => { | |
663 | minsize => 2048, | |
664 | name => 'RSA', | |
665 | }, | |
666 | 'id-ecPublicKey' => { | |
667 | minsize => 224, | |
668 | name => 'ECC', | |
669 | }, | |
670 | }; | |
671 | ||
672 | my $certs_check_failed = 0; | |
673 | foreach my $cert (@$certs) { | |
674 | my ($type, $size, $fn) = $cert->@{qw(public-key-type public-key-bits filename)}; | |
675 | ||
676 | if (!defined($type) || !defined($size)) { | |
677 | log_warn("'$fn': cannot check certificate, failed to get it's type or size!"); | |
678 | } | |
679 | ||
680 | my $check = $certs_check->{$type}; | |
681 | if (!defined($check)) { | |
682 | log_warn("'$fn': certificate's public key type '$type' unknown, check Debian Busters release notes"); | |
683 | next; | |
684 | } | |
685 | ||
686 | if ($size < $check->{minsize}) { | |
687 | log_fail("'$fn', certificate's $check->{name} public key size is less than 2048 bit"); | |
688 | $certs_check_failed = 1; | |
689 | } else { | |
690 | log_pass("Certificate '$fn' passed Debian Busters security level for TLS connections ($size >= 2048)"); | |
691 | } | |
692 | } | |
e522a2db FE |
693 | |
694 | check_backup_retention_settings(); | |
5df8b555 | 695 | check_cifs_credential_location(); |
4177a14d TL |
696 | } |
697 | ||
698 | __PACKAGE__->register_method ({ | |
699 | name => 'checklist', | |
700 | path => 'checklist', | |
701 | method => 'GET', | |
702 | description => 'Check (pre-/post-)upgrade conditions.', | |
703 | parameters => { | |
704 | additionalProperties => 0, | |
705 | properties => { | |
706 | }, | |
707 | }, | |
708 | returns => { type => 'null' }, | |
709 | code => sub { | |
710 | my ($param) = @_; | |
711 | ||
712 | check_pve_packages(); | |
713 | check_cluster_corosync(); | |
714 | check_ceph(); | |
715 | check_storage_health(); | |
716 | check_misc(); | |
717 | ||
718 | print_header("SUMMARY"); | |
719 | ||
720 | my $total = 0; | |
721 | $total += $_ for values %$counters; | |
722 | ||
723 | print "TOTAL: $total\n"; | |
724 | print colored("PASSED: $counters->{pass}\n", 'green'); | |
725 | print "SKIPPED: $counters->{skip}\n"; | |
726 | print colored("WARNINGS: $counters->{warn}\n", 'yellow'); | |
727 | print colored("FAILURES: $counters->{fail}\n", 'red'); | |
728 | ||
729 | if ($counters->{warn} > 0 || $counters->{fail} > 0) { | |
730 | my $color = $counters->{fail} > 0 ? 'red' : 'yellow'; | |
731 | print colored("\nATTENTION: Please check the output for detailed information!\n", $color); | |
732 | print colored("Try to solve the problems one at a time and then run this checklist tool again.\n", $color) if $counters->{fail} > 0; | |
733 | } | |
734 | ||
735 | return undef; | |
736 | }}); | |
737 | ||
738 | our $cmddef = [ __PACKAGE__, 'checklist', [], {}]; | |
739 | ||
740 | # for now drop all unknown params and just check | |
741 | @ARGV = (); | |
742 | ||
743 | 1; |