Enumerate all users in LDAP with PHP - php

I'd like to create a php script that runs as a daily cron. What I'd like to do is enumerate through all users within an Active Directory, extract certain fields from each entry, and use this information to update fields within a MySQL database.
Basically what I want to to do is sync up certain user information between Active Directory and a MySQL table.
The problem I have is that the sizelimit on the Active Directory server is often set at 1000 entries per search result. I had hoped that the php function "ldap_next_entry" would get around this by only fetching one entry at a time, but before you can call "ldap_next_entry", you first have to call "ldap_search", which can trigger the SizeLimit exceeded error.
Is there any way besides removing the sizelimit from the server? Can I somehow get "pages" of results?
BTW - I am currently not using any 3rd party libraries or code. Just PHPs ldap methods. Although, I am certainly open to using a library if that will help.

I've been struck by the same problem while developing Zend_Ldap for the Zend Framework. I'll try to explain what the real problem is, but to make it short: until PHP 5.4, it wasn't possible to use paged results from an Active Directory with an unpatched PHP (ext/ldap) version due to limitations in exactly this extension.
Let's try to unravel the whole thing... Microsoft Active Directory uses a so called server control to accomplish server-side result paging. This control ist described in RFC 2696 "LDAP Control Extension for Simple Paged Results Manipulation" .
ext/php offers an access to LDAP control extensions via its ldap_set_option() and the LDAP_OPT_SERVER_CONTROLS and LDAP_OPT_CLIENT_CONTROLS option respectively. To set the paged control you do need the control-oid, which is 1.2.840.113556.1.4.319, and we need to know how to encode the control-value (this is described in the RFC). The value is an octet string wrapping the BER-encoded version of the following SEQUENCE (copied from the RFC):
realSearchControlValue ::= SEQUENCE {
size INTEGER (0..maxInt),
-- requested page size from client
-- result set size estimate from server
cookie OCTET STRING
}
So we can set the appropriate server control prior to executing the LDAP query:
$pageSize = 100;
$pageControl = array(
'oid' => '1.2.840.113556.1.4.319', // the control-oid
'iscritical' => true, // the operation should fail if the server is not able to support this control
'value' => sprintf ("%c%c%c%c%c%c%c", 48, 5, 2, 1, $pageSize, 4, 0) // the required BER-encoded control-value
);
This allows us to send a paged query to the LDAP/AD server. But how do we know if there are more pages to follow and how do we specify with which control-value we have to send our next query?
This is where we're getting stuck... The server responds with a result set that includes the required paging information but PHP lacks a method to retrieve exactly this information from the result set. PHP provides a wrapper for the LDAP API function ldap_parse_result() but the required last parameter serverctrlsp is not exposed to the PHP function, so there is no way to retrieve the required information. A bug report has been filed for this issue but there has been no response since 2005. If the ldap_parse_result() function provided the required parameter, using paged results would work like
$l = ldap_connect('somehost.mydomain.com');
$pageSize = 100;
$pageControl = array(
'oid' => '1.2.840.113556.1.4.319',
'iscritical' => true,
'value' => sprintf ("%c%c%c%c%c%c%c", 48, 5, 2, 1, $pageSize, 4, 0)
);
$controls = array($pageControl);
ldap_set_option($l, LDAP_OPT_PROTOCOL_VERSION, 3);
ldap_bind($l, 'CN=bind-user,OU=my-users,DC=mydomain,DC=com', 'bind-user-password');
$continue = true;
while ($continue) {
ldap_set_option($l, LDAP_OPT_SERVER_CONTROLS, $controls);
$sr = ldap_search($l, 'OU=some-ou,DC=mydomain,DC=com', 'cn=*', array('sAMAccountName'), null, null, null, null);
ldap_parse_result ($l, $sr, $errcode, $matcheddn, $errmsg, $referrals, $serverctrls); // (*)
if (isset($serverctrls)) {
foreach ($serverctrls as $i) {
if ($i["oid"] == '1.2.840.113556.1.4.319') {
$i["value"]{8} = chr($pageSize);
$i["iscritical"] = true;
$controls = array($i);
break;
}
}
}
$info = ldap_get_entries($l, $sr);
if ($info["count"] < $pageSize) {
$continue = false;
}
for ($entry = ldap_first_entry($l, $sr); $entry != false; $entry = ldap_next_entry($l, $entry)) {
$dn = ldap_get_dn($l, $entry);
}
}
As you see there is a single line of code (*) that renders the whole thing useless. On my way though the sparse information on this subject I found a patch against the PHP 4.3.10 ext/ldap by IƱaki Arenaza but neither did I try it nor do I know if the patch can be applied on a PHP5 ext/ldap. The patch extends ldap_parse_result() to expose the 7th parameter to PHP:
--- ldap.c 2004-06-01 23:05:33.000000000 +0200
+++ /usr/src/php4/php4-4.3.10/ext/ldap/ldap.c 2005-09-03 17:02:03.000000000 +0200
## -74,7 +74,7 ##
ZEND_DECLARE_MODULE_GLOBALS(ldap)
static unsigned char third_argument_force_ref[] = { 3, BYREF_NONE, BYREF_NONE, BYREF_FORCE };
-static unsigned char arg3to6of6_force_ref[] = { 6, BYREF_NONE, BYREF_NONE, BYREF_FORCE, BYREF_FORCE, BYREF_FORCE, BYREF_FORCE };
+static unsigned char arg3to7of7_force_ref[] = { 7, BYREF_NONE, BYREF_NONE, BYREF_FORCE, BYREF_FORCE, BYREF_FORCE, BYREF_FORCE, BYREF_FORCE };
static int le_link, le_result, le_result_entry, le_ber_entry;
## -124,7 +124,7 ##
#if ( LDAP_API_VERSION > 2000 ) || HAVE_NSLDAP
PHP_FE(ldap_get_option, third_argument_force_ref)
PHP_FE(ldap_set_option, NULL)
- PHP_FE(ldap_parse_result, arg3to6of6_force_ref)
+ PHP_FE(ldap_parse_result, arg3to7of7_force_ref)
PHP_FE(ldap_first_reference, NULL)
PHP_FE(ldap_next_reference, NULL)
#ifdef HAVE_LDAP_PARSE_REFERENCE
## -1775,14 +1775,15 ##
Extract information from result */
PHP_FUNCTION(ldap_parse_result)
{
- pval **link, **result, **errcode, **matcheddn, **errmsg, **referrals;
+ pval **link, **result, **errcode, **matcheddn, **errmsg, **referrals, **serverctrls;
ldap_linkdata *ld;
LDAPMessage *ldap_result;
+ LDAPControl **lserverctrls, **ctrlp, *ctrl;
char **lreferrals, **refp;
char *lmatcheddn, *lerrmsg;
int rc, lerrcode, myargcount = ZEND_NUM_ARGS();
- if (myargcount 6 || zend_get_parameters_ex(myargcount, &link, &result, &errcode, &matcheddn, &errmsg, &referrals) == FAILURE) {
+ if (myargcount 7 || zend_get_parameters_ex(myargcount, &link, &result, &errcode, &matcheddn, &errmsg, &referrals, &serverctrls) == FAILURE) {
WRONG_PARAM_COUNT;
}
## -1793,7 +1794,7 ##
myargcount > 3 ? &lmatcheddn : NULL,
myargcount > 4 ? &lerrmsg : NULL,
myargcount > 5 ? &lreferrals : NULL,
- NULL /* &serverctrls */,
+ myargcount > 6 ? &lserverctrls : NULL,
0 );
if (rc != LDAP_SUCCESS ) {
php_error(E_WARNING, "%s(): Unable to parse result: %s", get_active_function_name(TSRMLS_C), ldap_err2string(rc));
## -1805,6 +1806,29 ##
/* Reverse -> fall through */
switch(myargcount) {
+ case 7 :
+ zval_dtor(*serverctrls);
+
+ if (lserverctrls != NULL) {
+ array_init(*serverctrls);
+ ctrlp = lserverctrls;
+
+ while (*ctrlp != NULL) {
+ zval *ctrl_array;
+
+ ctrl = *ctrlp;
+ MAKE_STD_ZVAL(ctrl_array);
+ array_init(ctrl_array);
+
+ add_assoc_string(ctrl_array, "oid", ctrl->ldctl_oid,1);
+ add_assoc_bool(ctrl_array, "iscritical", ctrl->ldctl_iscritical);
+ add_assoc_stringl(ctrl_array, "value", ctrl->ldctl_value.bv_val,
+ ctrl->ldctl_value.bv_len,1);
+ add_next_index_zval (*serverctrls, ctrl_array);
+ ctrlp++;
+ }
+ ldap_controls_free (lserverctrls);
+ }
case 6 :
zval_dtor(*referrals);
if (array_init(*referrals) == FAILURE) {
Actually the only option left would be to change the Active Directory configuration and raise the maximum result limit. The relevant option is called MaxPageSize and can be altered by using ntdsutil.exe - please see "How to view and set LDAP policy in Active Directory by using Ntdsutil.exe".
EDIT (reference to COM):
Or you can go the other way round and use the COM-approach via ADODB as suggested in the link provided by eykanal.

Support for paged results was added in PHP 5.4.
See ldap_control_paged_result for more details.

This isn't a full answer, but this guy was able to do it. I don't understand what he did, though.
By the way, a partial answer is that you CAN get "pages" of results. From the documentation:
resource ldap_search ( resource $link_identifier , string $base_dn ,
string $filter [, array $attributes [, int $attrsonly [, int $sizelimit [,
int $timelimit [, int $deref ]]]]] )
...
sizelimit Enables you to limit the count of entries fetched. Setting this to 0 means no limit.
Note: This parameter can NOT override server-side preset sizelimit.
You can set it lower though. Some directory server hosts will be
configured to return no more than a preset number of entries. If this
occurs, the server will indicate that it has only returned a partial
results set. This also occurs if you use this parameter to limit the
count of fetched entries.
I don't know how to specify that you want to search STARTING from a certain position, though. I.e., after you get your first 1000, I don't know how to specify that now you need the next 1000. Hopefully someone else can help you there :)

Here's an alternative (which works pre PHP 5.4). If you have 10,000 records you need to get but your AD server only returns 5,000 per page:
$ldapSearch = ldap_search($ldapResource, $basedn, $filter, array('member;range=0-4999'));
$ldapResults = ldap_get_entries($dn, $ldapSearch);
$members = $ldapResults[0]['member;range=0-4999'];
$ldapSearch = ldap_search($ldapResource, $basedn, $filter, array('member;range=5000-10000'));
$ldapResults = ldap_get_entries($dn, $ldapSearch);
$members = array_merge($members, $ldapResults[0]['member;range=5000-*']);

I was able to get around the size limitation using ldap_control_paged_result
ldap_control_paged_result is used to Enable LDAP pagination by sending the pagination control. The below function worked perfectly in my case.
function retrieves_users($conn)
{
$dn = 'ou=,dc=,dc=';
$filter = "(&(objectClass=user)(objectCategory=person)(sn=*))";
$justthese = array();
// enable pagination with a page size of 100.
$pageSize = 100;
$cookie = '';
do {
ldap_control_paged_result($conn, $pageSize, true, $cookie);
$result = ldap_search($conn, $dn, $filter, $justthese);
$entries = ldap_get_entries($conn, $result);
if(!empty($entries)){
for ($i = 0; $i < $entries["count"]; $i++) {
$data['usersLdap'][] = array(
'name' => $entries[$i]["cn"][0],
'username' => $entries[$i]["userprincipalname"][0]
);
}
}
ldap_control_paged_result_response($conn, $result, $cookie);
} while($cookie !== null && $cookie != '');
return $data;
}

Related

Php Recursive Function Infinite loop Error

I am having a php recursive function to calculate nearest sale price. but i don't know why its run infinite time and throw error of maximum execution.
Its look like below:
function getamazonsaleper($portal)
{
$cp = floatval($this->input->post('cp')); //user provided inputs
$sp = floatval($this->input->post('sp')); //user provided input
$gst = floatval($this->input->post('gst')); //user provided input
$rfsp = floatval($this->input->post('rfsp')); //user provided input
$mcp = (int)($this->input->post('mcp')); //user provided input
$weight = floatval($this->input->post('weight')); //user provided input
$output = $this->getsalepercent($cp,$sp,$gst,$rfsp,$mcp,$weight,$portal);
return $output;
}
function getsalepercent($cp,$sp,$gst,$rfsp,$mcp,$weight,$portal) //recursive funtion
{
$spcost = ((($sp/100)*$cp));
$gstamount= (($spcost/(100+$gst))*$gst);
$rfspamount= ($spcost*($rfsp/100));
$mcpamount= ($cp*($mcp/100));
$fixedfee=$this->getfixedfee($portal,$spcost);
$weightfee=$this->getweightprice($portal,$weight);
$totalcost=$fixedfee+$weightfee+$rfspamount;
$gstinput=($totalcost*(18/100));
$remittances = $spcost-$totalcost-$gstinput;
$actualprofit= $remittances-$cp-$gstamount+$gstinput;
$actualprofitpercent = ($actualprofit/$cp)*100;
if( $actualprofitpercent >= $mcp)
{
return $sp;
}elseif($actualprofitpercent < $mcp)
{
$newsp = (int)($sp+10) ;
$this->getsalepercent($cp,$newsp,$gst,$rfsp,$mcp,$weight,$portal);
}
}
Can anybody tell me how can resolve this issue? Thanks in advance.
Edited :
Perameters
$cp=100;
$sp=200;
$mcp=20;
$weight=0.5;
$gst=28;
$rfsp=6.5;
First a couple of side notes:
- the way you use $gstinput it cancels itself out when you calculate $actualprofit (it's -$gstinput in $remittances which gets added to +$gstinput).
- $mcpamount seems to go completely unused in the code... I thought for a second you might vahe simply confused vars when doing the comparison, but of course for $cp = 100 it makes no difference.
Even so when I made a few calculations using the example values you gave for $sp = 200 (and growing by 10), I got:
Value of $actualprofit, which for $cp = 100 is also the value of $actualprofitpercent...
for $sp = 200:
43.25 - $fixedfee - $weightfee
for $sp = 210:
50.4125 - $fixedfee - $weightfee
for $sp = 220:
57.575 - $fixedfee - $weightfee
so for each $sp = $sp + 10 recursion the value of $actualprofitpercent (without taking into account $fixedfee and $weightfee) seems to grow by 7.1625.
The value of $weightfee should stay the same, but the value of $fixedfee depends on the value of $sp... Could it be that at each recursion getfixedfee() returns a value which grows faster than 7.1625?

php RRD graph separation

I am trying to create RRD graphs with the help of PHP in order to keep track of the inoctets,outoctets and counter of a server.
So far the script is operating as expected but my problems comes when I am trying to produce 2 or more separate graphs. I am trying to produce (hourly, weekly , etc) graphs. I thought by creating a loop would solve my problem, since I have split the RRA in hours and days. Unfortunately I end up having 2 graphs that updating simultaneously as expected but plotting the same thing. Has any one encounter similar problem? I have applied the same program in perl with RRD::Simple,where is extremely easy and everything is adjusted almost automatically.
I have supplied under a working example of my code with the minimum possible data because the code is a bit long:
<?php
$file = "snmp-2";
$rrdFile = dirname(__FILE__) . "/snmp-2.rrd";
$in = "ifInOctets";
$out = "ifOutOctets";
$count = "sysUpTime";
$step = 5;
$rounds = 1;
$output = array("Hourly","Daily");
while (1) {
sleep (6);
$options = array(
"--start","now -15s", // Now -10 seconds (default)
"--step", "".$step."",
"DS:".$in.":GAUGE:10:U:U",
"DS:".$out.":GAUGE:10:U:U",
"DS:".$count.":ABSOLUTE:10:0:4294967295",
"RRA:MIN:0.5:12:60",
"RRA:MAX:0.5:12:60",
"RRA:LAST:0.5:12:60",
"RRA:AVERAGE:0.5:12:60",
"RRA:MIN:0.5:300:60",
"RRA:MAX:0.5:300:60",
"RRA:LAST:0.5:300:60",
"RRA:AVERAGE:0.5:300:60",
);
if ( !isset( $create ) ) {
$create = rrd_create(
"".$rrdFile."",
$options
);
if ( $create === FALSE ) {
echo "Creation error: ".rrd_error()."\n";
}
}
$t = time();
$ifInOctets = rand(0, 4294967295);
$ifOutOctets = rand(0, 4294967295);
$sysUpTime = rand(0, 4294967295);
$update = rrd_update(
"".$rrdFile."",
array(
"".$t.":".$ifInOctets.":".$ifOutOctets.":".$sysUpTime.""
)
);
if ($update === FALSE) {
echo "Update error: ".rrd_error()."\n";
}
$start = $t - ($step * $rounds);
foreach ($output as $test) {
$final = array(
"--start","".$start." -15s",
"--end", "".$t."",
"--step","".$step."",
"--title=".$file." RRD::Graph",
"--vertical-label=Byte(s)/sec",
"--right-axis-label=latency(min.)",
"--alt-y-grid", "--rigid",
"--width", "800", "--height", "500",
"--lower-limit=0",
"--alt-autoscale-max",
"--no-gridfit",
"--slope-mode",
"DEF:".$in."_def=".$file.".rrd:".$in.":AVERAGE",
"DEF:".$out."_def=".$file.".rrd:".$out.":AVERAGE",
"DEF:".$count."_def=".$file.".rrd:".$count.":AVERAGE",
"CDEF:inbytes=".$in."_def,8,/",
"CDEF:outbytes=".$out."_def,8,/",
"CDEF:counter=".$count."_def,8,/",
"COMMENT:\\n",
"LINE2:".$in."_def#FF0000:".$in."",
"COMMENT:\\n",
"LINE2:".$out."_def#0000FF:".$out."",
"COMMENT:\\n",
"LINE2:".$count."_def#FFFF00:".$count."",
);
$outputPngFile = rrd_graph(
"".$test.".png",
$final
);
if ($outputPngFile === FALSE) {
echo "<b>Graph error: </b>".rrd_error()."\n";
}
} /* End of foreach function */
$debug = rrd_lastupdate (
"".$rrdFile.""
);
if ($debug === FALSE) {
echo "<b>Graph result error: </b>".rrd_error()."\n";
}
var_dump ($debug);
$rounds++;
} /* End of while loop */
?>
A couple of issues.
Firstly, your definition of the RRD has a step of 5seconds and RRAs with steps of 12x5s=1min and 300x5s=25min. They also have a length of only 60 rows, so 1hr and 25hr respectively. You'll never get a weekly graph this way! You need to add more rows; also the step seems rather short, and you might need a smaller-step RRA for hourly graphs and a larger-step one for weekly graphs.
Secondly, it is not clear how you're calling the graph function. You seem to be specifying:
"--start","".$start." -15s",
"--end", "".$t."",
"--step","".$step."",
... which would force it to use the 5s interval (unavailable, so the 1min one would always get used) and for the graph to be only for the time window from the start to the last update, not a 'hourly' or 'daily' as you were asking.
Note that the RRA you have defined do not define the time window of the graph you are asking for. Also, just because you have more than one RRA defined, it doesnt mean you'll get more than one graph unless oyu call the graph function twice with different arguments.
If you want a daily graph, use
"--start","end - 1 hour",
"--end",$t,
Do not specify a step as the most appropriate available will be used anyway. For a daily graph, use
"--start","end - 1 day"
"--end",$t,
Similarly, no need to specify a step.
Hopefully this will make it a little clearer. Most of the RRD graph options have sensible defaults, and RRDTool is pretty good at picking the correct RRA to use based on the graph size, time window, and DEF statements.

Insert performance of node-mongodb-native

I'm testing performance of Node.js with MongoDB. I know each of these is fine independent of the other, but I'm trying a handful of tests to get a feel for them. I ran across this issue and I'm having trouble determining the source.
The Problem
I'm trying to insert 1,000,000 records in a single Node.js program. It absolutely crawls. We're talking 20 minute execution time. This occurs whether it's my Mac or CentOS, although the behavior is marginally different between the two. It does eventually complete.
The effect is similar to swapping, although it's not (memory never exceeds 2 GB). There are only 3 connections open to MongoDB, and most of the time there's no data being inserted. It appears to be doing a lot of context switching, and the Node.js CPU core is maxed out.
The effect is similar to the one mentioned in this thread.
I try the same using PHP and it finishes in 2-3 minutes. No drama.
Why?
Possible Causes
I currently believe this is either a Node.js socket issue, something going on with libev behind the scenes, or some other node-mongodb-native issue. I may be totally wrong, so I'm looking for a little guidance here.
As for other Node.js MongoDB adapters, I have tried Mongolian and it appears to queue documents in order to batch insert them, and it ends up running out of memory. So that's out. (Side note: I have no idea why on this, either, since it doesn't even come close to my 16 GB box limit--but I haven't bothered investigating much further on that.)
I should probably mention that I did in fact test a master/worker cluster with 4 workers (on a quad-core machine) and it finished in 2-3 minutes.
The Code
Here's my Node.js CoffeeScript program:
mongodb = require "mongodb"
microtime = require "microtime"
crypto = require "crypto"
times = 1000000
server = new mongodb.Server "127.0.0.1", 27017
db = mongodb.Db "test", server
db.open (error, client) ->
throw error if error?
collection = mongodb.Collection client, "foo"
for i in [0...times]
console.log "Inserting #{i}..." if i % 100000 == 0
hash = crypto.createHash "sha1"
hash.update "" + microtime.now() + (Math.random() * 255 | 0)
key = hash.digest "hex"
doc =
key: key,
foo1: 1000,
foo2: 1000,
foo3: 1000,
bar1: 2000,
bar2: 2000,
bar3: 2000,
baz1: 3000,
baz2: 3000,
baz3: 3000
collection.insert doc, safe: true, (error, response) ->
console.log error.message if error
And here's the roughly equivalent PHP program:
<?php
$mongo = new Mongo();
$collection = $mongo->test->foo;
$times = 1000000;
for ($i = 0; $i < $times; $i++) {
if ($i % 100000 == 0) {
print "Inserting $i...\n";
}
$doc = array(
"key" => sha1(microtime(true) + rand(0, 255)),
"foo1" => 1000,
"foo2" => 1000,
"foo3" => 1000,
"bar1" => 2000,
"bar2" => 2000,
"bar3" => 2000,
"baz1" => 3000,
"baz2" => 3000,
"baz3" => 3000
);
try {
$collection->insert($doc, array("safe" => true));
} catch (MongoCursorException $e) {
print $e->getMessage() . "\n";
}
}
It sounds like you're running into the default heap limit in V8. I wrote a blog post about removing this limitation.
The garbage collector is probably going crazy and chewing on CPU, since it will constantly execute until you're under the 1.4GB limit.
What happens if you explicitly return a value at the end of the db.open callback function? Your generated javascript code is pushing all of your collection.insert returns onto a big "_results" array, which would get slower and slower, I imagine.
db.open(function(error, client) {
var collection, doc, hash, i, key, _i, _results;
if (error != null) {
throw error;
}
collection = mongodb.Collection(client, "foo");
_results = [];
for (i = _i = 0; 0 <= times ? _i < times : _i > times; i = 0 <= times ? ++_i : --_i) {
...
_results.push(collection.insert(doc, {
safe: true
}, function(error, response) {
if (error) {
return console.log(error.message);
}
}));
}
return _results;
});
Try adding this at the end of your coffeescript:
collection.insert doc, safe: true, (error, response) ->
console.log error.message if error
return
*Update: * So, I actually tried to run your program, and noticed a few more issues:
The biggest problem is you're trying to spawn a million inserts in a synchronous fashion, which will really kill your RAM, and eventually stop inserting (at least, it did for me). I killed it at 800MB RAM or so.
You need to change the way you're calling collection.insert() so that it works asynchronously.
I rewrote it like so, breaking out a couple of functions for clarity:
mongodb = require "mongodb"
microtime = require "microtime"
crypto = require "crypto"
gen = () ->
hash = crypto.createHash "sha1"
hash.update "" + microtime.now() + (Math.random() * 255 | 0)
key = hash.digest "hex"
key: key,
foo1: 1000,
foo2: 1000,
foo3: 1000,
bar1: 2000,
bar2: 2000,
bar3: 2000,
baz1: 3000,
baz2: 3000,
baz3: 3000
times = 1000000
i = times
insertDocs = (collection) ->
collection.insert gen(), {safe:true}, () ->
console.log "Inserting #{times-i}..." if i % 100000 == 0
if --i > 0
insertDocs(collection)
else
process.exit 0
return
server = new mongodb.Server "127.0.0.1", 27017
db = mongodb.Db "test", server
db.open (error, db) ->
throw error if error?
db.collection "foo", (err, collection) ->
insertDocs(collection)
return
return
Which finished in ~3 minutes:
wfreeman$ time coffee mongotest.coffee
Inserting 0...
Inserting 100000...
Inserting 200000...
Inserting 300000...
Inserting 400000...
Inserting 500000...
Inserting 600000...
Inserting 700000...
Inserting 800000...
Inserting 900000...
real 3m31.991s
user 1m55.211s
sys 0m23.420s
Also, it has the side benefit of using <100MB of RAM, 70% CPU on node, and 40% CPU on mongod (on a 2 core box, so it looks like it wasn't maxing out the CPU).

Predis Protocol error: invalid bulk length

I am using predis and everything was great until I started getting this error:
ERR Protocol error: invalid bulk length
I am not sure why I am getting it. The error is in this file: Predis/Network/StreamConnection.php in this method:
public function writeCommand(ICommand $command) {
$commandId = $command->getId();
$arguments = $command->getArguments();
$cmdlen = strlen($commandId);
$reqlen = count($arguments) + 1;
$buffer = "*{$reqlen}\r\n\${$cmdlen}\r\n{$commandId}\r\n";
for ($i = 0; $i < $reqlen - 1; $i++) {
$argument = $arguments[$i];
$arglen = strlen($argument);
$buffer .= "\${$arglen}\r\n{$argument}\r\n";
}
$this->writeBytes($buffer);
}
It fails when it tries to do an strlen() on an array.
Here is the code that is causing this to fail:
$ids = array(1, 2, 3);
$predis = new Predis\Client();
$predis->set('testerKey', $ids);
Am I not allowed to set an array? Of course I can set an array. The only thing I changed was I make my files UTF-8 so maybe that screwed something up?
Any help would be appreciated.
I found the problem and a solution. Coming from memcached where it will serialize the array automatically this is not the same in PRedis. PRedis will never serialize anything when performing a set or get.
https://github.com/nrk/predis/issues/29
You have to use mset.
With the set command, Predis is looking for an array with only 2 variables (to set the key => hash). Do set 3 keys, you have to use mset.
To do what you seem to be trying to do:
$ids = array(1 => 'id-1', 2 => 'id-2', 3 => 'id-3');
$predis = new Predis\Client();
$predis->mset('testerKey', $ids);

SNMP Mapping Mac Address to Switch Port

I work at a datacenter and I'm in the process of writing a php tool that maps all of our devices and can tell us if what is out there is what is being billed for.
It first pulls a huge list of macs and their ips from both of the cores into a temp table. Then, it loops through all of the racks* and attempts to find which port that mac belongs to. Since there is no golden command (cue lightbulb over your head), I have to:
Create a multi-array with the port as the key and the ifindex for the value.
Replace the ifindex with with a bridge ID.
Replace the bridge ID with the mac hash.
Repalce the mac hash with the actual mac
Lastly, it takes the mac, ips, and port and populates the master table.
The problem is step one. 1.3.6.1.2.1.31.1.1.1.1 works on most of the switches but a few of the foundrys do not work. 1.3.6.1.4.1.1991.1.1.3.3.1.1.38 kinda comes close to what I'm looking for but im not entirely comfortable it's what I'm looking for. I was able to find the specific device models under foundry > products > registration, but there aren't any MIBs under that folder. So my questions are:
Is there a foundry specific string that returns ports and macs? ifindexes would also work.
How do I go about using device specific MIBs (enterprises.foundry.products.registration.snFWSXFamily)?
Any direction on this would be great.
-Justin
*= rack models: cisco 2900xl, foundry FI4802 + variants
You can do this (tested on HP Procurve) :
From your linux server :
$ snmpwalk -v 1 -c public xxx.xxx.xxx.xxx 1.3.6.1.2.1.17.4.3.1.2 | grep
"INTEGER: 11"
(port number 11)
Will return :
SNMPv2-SMI::mib-2.17.4.3.1.2.44.118.138.64.143.95 = INTEGER: 11
SNMPv2-SMI::mib-2.17.4.3.1.2.56.170.60.108.174.57 = INTEGER: 11
SNMPv2-SMI::mib-2.17.4.3.1.2.104.181.153.172.54.237 = INTEGER: 11
SNMPv2-SMI::mib-2.17.4.3.1.2.120.172.192.143.226.236 = INTEGER: 11
SNMPv2-SMI::mib-2.17.4.3.1.2.124.195.161.20.109.76 = INTEGER: 11
SNMPv2-SMI::mib-2.17.4.3.1.2.152.75.225.59.127.180 = INTEGER: 11
Then you can do this to find which Mac Address is connected :
$ snmpwalk -v 1 -c public xxx.xxx.xxx.xxx 1.3.6.1.2.1.17.4.3.1.1 |
grep "152.75.225.59.127.180"
Return mac address :
SNMPv2-SMI::mib-2.17.4.3.1.1.152.75.225.59.127.180 = Hex-STRING: 98
4B E1 3B 7F B4
You can make a script.sh to do this...
when I needed to discover MACs and some other info from my switches, I used 'snmpwalk' and 'snmpbulkwalk' commands to examine their SNMP data contents
for example:
snmpbulkwalk -v2c 192.168.30.40 -c public 1.3.6.1.2.1.31.1.1.1.1
outputs:
IF-MIB::ifName.1 = STRING: Gi0/1
IF-MIB::ifName.2 = STRING: Gi0/2
IF-MIB::ifName.3 = STRING: Gi0/3
IF-MIB::ifName.4 = STRING: Gi0/4
IF-MIB::ifName.5 = STRING: Gi0/5
IF-MIB::ifName.6 = STRING: Gi0/6
IF-MIB::ifName.7 = STRING: Gi0/7
IF-MIB::ifName.8 = STRING: Gi0/8
IF-MIB::ifName.9 = STRING: Gi0/9
IF-MIB::ifName.10 = STRING: Gi0/10
IF-MIB::ifName.11 = STRING: Gi0/11
IF-MIB::ifName.12 = STRING: Gi0/12
IF-MIB::ifName.13 = STRING: Nu0
IF-MIB::ifName.14 = STRING: Vl1
IF-MIB::ifName.15 = STRING: Vl2
IF-MIB::ifName.16 = STRING: Vl416
and
snmpbulkwalk -v2c 192.168.30.40 -c public 1.3.6.1.2
outputs A LOT of info among which you can look for your favorite MACs or anything
If any one you would like to do this programmatically from a Windows Server you can use the SnmpSharpNet library to accomplish the same thing. Here's an example that will create a list of all the MAC Addresses and Ports on a particular Dell Switch using the OID 1.3.6.1.2.1.17.7.1.2.2.1.2.1
using SnmpSharpNet;
List<KeyValuePair<string, string>> portList = new List<KeyValuePair<string, string>>();
IPAddress ip = IPAddress.Parse("192.168.0.2");
SnmpWalk(ip, "snmpcommunity", "1.3.6.1.2.1.17.7.1.2.2.1.2.1", "1");
//SNMPWALK the ports on a switch or stack of switches. Ports will be labeled SwitchNum/Stack Number/Port Numbers.
private void SnmpWalk(IPAddress ip, string snmpCommunity, string oid, string switchNum)
{
UdpTarget target = new UdpTarget(ip);
// SNMP community name
OctetString community = new OctetString(snmpCommunity);
// Define agent parameters class
AgentParameters param = new AgentParameters(community);
// Set SNMP version to 1
param.Version = SnmpVersion.Ver1;
// Define Oid that is the root of the MIB tree you wish to retrieve
Oid rootOid = new Oid(oid);
// This Oid represents last Oid returned by the SNMP agent
Oid lastOid = (Oid)rootOid.Clone();
// Pdu class used for all requests
Pdu pdu = new Pdu(PduType.GetNext);
// Loop through results
while (lastOid != null)
{
// When Pdu class is first constructed, RequestId is set to a random value
// that needs to be incremented on subsequent requests made using the
// same instance of the Pdu class.
if (pdu.RequestId != 0)
{
pdu.RequestId += 1;
}
// Clear Oids from the Pdu class.
pdu.VbList.Clear();
// Initialize request PDU with the last retrieved Oid
pdu.VbList.Add(lastOid);
// Make SNMP request
SnmpV1Packet result = (SnmpV1Packet)target.Request(pdu, param);
// You should catch exceptions in the Request if using in real application.
// If result is null then agent didn't reply or we couldn't parse the reply.
if (result != null)
{
// ErrorStatus other then 0 is an error returned by
// the Agent - see SnmpConstants for error definitions
if (result.Pdu.ErrorStatus != 0)
{
// agent reported an error with the request
Console.WriteLine("Error in SNMP reply. Error {0} index {1}",
result.Pdu.ErrorStatus,
result.Pdu.ErrorIndex);
lastOid = null;
break;
}
else
{
// Walk through returned variable bindings
foreach (Vb v in result.Pdu.VbList)
{
// Check that retrieved Oid is "child" of the root OID
if (rootOid.IsRootOf(v.Oid))
{
//Convert OID to MAC
string[] macs = v.Oid.ToString().Split('.');
string mac = "";
int counter = 0;
foreach (string chunk in macs)
{
if (counter >= macs.Length - 6)
{
mac += string.Format("{0:X2}", int.Parse(chunk));
}
counter += 1;
}
//Assumes a 48 port switch (52 actual). You need to know these values to correctly iterate through a stack of switches.
int dellSwitch = 1 + int.Parse(v.Value.ToString()) / 52;
int port = int.Parse(v.Value.ToString()) - (52 * (dellSwitch - 1));
KeyValuePair<string, string> Port = new KeyValuePair<string, string>(mac, switchNum + "/" + dellSwitch.ToString() + "/" + port.ToString());
portList.Add(Port);
//Exit Loop
lastOid = v.Oid;
}
else
{
//End of the requested MIB tree. Set lastOid to null and exit loop
lastOid = null;
}
}
}
}
else
{
Console.WriteLine("No response received from SNMP agent.");
}
}
target.Close();
}
There is an entire project utilizing this example that resolves this information back to Hostnames here

Categories