Operating SystemsI/O Software

Spooling: Simultaneous Peripheral Operations Online

LevelIntermediate

Duration60 mins

TopicI/O Software

2 / 5

Print Spooler: Architecture and Implementation

The Print Spooler: Where Theory Meets Practice

The print spooler is the most visible and widely-used implementation of spooling principles. Every time you print a document—whether from a word processor, web browser, or any application—a sophisticated subsystem springs into action, managing the journey from digital data to physical output.

Modern print spoolers are remarkably complex, handling diverse printer types (inkjet, laser, thermal, 3D), connection methods (USB, network, Bluetooth), document formats (PDF, PostScript, raw), and features (duplex, collation, stapling). Understanding this system reveals both the practical application of spooling concepts and the engineering challenges of real-world I/O management.

What You Will Learn

This page covers print spooler architecture in depth: CUPS (Common UNIX Printing System), the IPP protocol, print filters and backends, job processing pipelines, and practical administration. You'll understand the complete journey from application print call to ink on paper.

CUPS Architecture

CUPS (Common UNIX Printing System) is the standard printing system on Linux, macOS, and many UNIX variants. Developed by Apple (acquired from Easy Software Products in 2007), CUPS provides a modern, standards-based printing architecture.

Core Components:

cupsd - The scheduler daemon that manages print queues and jobs
cups-browsed - Discovers network printers and creates local queues
Filters - Convert document formats for specific printers
Backends - Handle communication with physical devices
PPD files - Describe printer capabilities and options

Design Philosophy:

CUPS uses HTTP/IPP (Internet Printing Protocol) for all communication, even locally. This unified approach means the same protocol works for local USB printers and remote network printers.

Converting Mermaid diagram...

cups_job_submission.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
/* CUPS Print Job Submission Example */
#include <cups/cups.h>
#include <stdio.h>
 
int submit_print_job(const char *filename, const char *printer) {
    cups_option_t *options = NULL;
    int num_options = 0;
    
    /* Configure print options */
    num_options = cupsAddOption("copies", "2", num_options, &options);
    num_options = cupsAddOption("media", "Letter", num_options, &options);
    num_options = cupsAddOption("sides", "two-sided-long-edge", 
                                 num_options, &options);
    num_options = cupsAddOption("print-quality", "5", num_options, &options);
    
    /* Submit job - returns immediately after spooling */
    int job_id = cupsPrintFile(printer, filename, "Document", 
                                num_options, options);
    
    cupsFreeOptions(num_options, options);
    
    if (job_id == 0) {
        fprintf(stderr, "Error: %s
", cupsLastErrorString());
        return -1;
    }
    
    printf("Job %d submitted to %s
", job_id, printer);
    return job_id;
}
 
/* Monitor job status */
void monitor_job(int job_id) {
    cups_job_t *jobs;
    int num_jobs = cupsGetJobs(&jobs, NULL, 0, CUPS_WHICHJOBS_ALL);
    
    for (int i = 0; i < num_jobs; i++) {
        if (jobs[i].id == job_id) {
            printf("Job %d: %s - %s
", 
                   jobs[i].id, jobs[i].title,
                   jobs[i].state == IPP_JSTATE_PENDING ? "Pending" :
                   jobs[i].state == IPP_JSTATE_PROCESSING ? "Processing" :
                   jobs[i].state == IPP_JSTATE_COMPLETED ? "Completed" :
                   "Unknown");
            break;
        }
    }
    cupsFreeJobs(num_jobs, jobs);
}

Internet Printing Protocol (IPP)

IPP is the foundational protocol of modern network printing. An IETF standard (RFC 8011), IPP uses HTTP as transport and defines operations for print job submission, management, and status queries.

Key IPP Operations:

Operation	Purpose
Print-Job	Submit a new print job
Validate-Job	Check if job would be accepted
Get-Jobs	List jobs in a queue
Get-Job-Attributes	Query specific job status
Cancel-Job	Cancel a pending/processing job
Get-Printer-Attributes	Query printer capabilities
Hold-Job / Release-Job	Manage job scheduling

IPP Advantages:

Firewall-friendly: Uses standard HTTP/HTTPS ports
Secure: Supports TLS encryption and authentication
Feature-rich: Standardized attributes for all print options
Bidirectional: Clients can query status and receive events

IPP vs Legacy Protocols
Feature	IPP	LPD/LPR	Raw (Port 9100)
Transport	HTTP/HTTPS	TCP	TCP
Security	TLS, Authentication	None	None
Bidirectional	Yes	Limited	No
Job Status	Rich attributes	Basic	None
Standard	IETF RFC 8011	RFC 1179	De facto
Port	631	515	9100

The Filter Chain: Document Transformation

CUPS transforms documents through a filter chain—a pipeline of programs that convert the input format to the printer's native language.

Example Filter Chain:

PDF → pstops → pdftops → foomatic-rip → hp (backend)

Common Filters:

pdftops/pdftoraster: Convert PDF to PostScript or raster
pstops: Impose pages (n-up, booklet), select page ranges
foomatic-rip: Generic driver wrapper using PPD options
imagetopdf: Convert images to PDF for further processing

Filter Selection:

CUPS automatically determines the filter chain based on:

Input MIME type (detected or specified)
Printer's supported formats (from PPD)
Available filters in /usr/lib/cups/filter/

The scheduler traces a path through a MIME conversion graph to find the shortest chain from input to printer-native format.

simple_filter.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
/* Simple CUPS Filter Example
 * Filters read from stdin or file, write to stdout
 * Arguments: job-id user title copies options [filename]
 */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
 
int main(int argc, char *argv[]) {
    FILE *input;
    char buffer[8192];
    size_t bytes;
    
    /* Open input: file argument or stdin */
    if (argc > 6) {
        input = fopen(argv[6], "rb");
        if (!input) {
            fprintf(stderr, "ERROR: Cannot open %s
", argv[6]);
            return 1;
        }
    } else {
        input = stdin;
    }
    
    /* Log to stderr (appears in error_log) */
    fprintf(stderr, "DEBUG: Processing job %s for %s
", 
            argv[1], argv[2]);
    
    /* Process: this example just passes through */
    /* Real filter would transform the data */
    while ((bytes = fread(buffer, 1, sizeof(buffer), input)) > 0) {
        /* Apply transformation here */
        fwrite(buffer, 1, bytes, stdout);
    }
    
    if (argc > 6) fclose(input);
    return 0;
}

Debugging Filter Chains

Use cupsfilter -p printer.ppd -m printer/type input.pdf > output.prn to manually run the filter chain. Check /var/log/cups/error_log for detailed filter execution logs. Set LogLevel debug in cupsd.conf for maximum verbosity.

Backends: Device Communication

Backends are the final stage—programs that transmit filtered job data to physical devices. Each backend handles a specific connection type.

Standard CUPS Backends:

CUPS Backend Types
Backend	URI Scheme	Purpose
usb	usb://make/model	USB-connected printers
socket	socket://host:9100	Raw TCP connection
ipp/ipps	ipp://host/path	IPP network printers
lpd	lpd://host/queue	Legacy LPD printers
smb	smb://host/share	Windows shared printers
serial	serial:/dev/ttyS0	Serial port printers
cups-pdf	cups-pdf:/	Virtual PDF printer

Backend Responsibilities:

Device discovery: Report available devices via backend --list
Connection management: Establish and maintain device connection
Error handling: Detect and report device errors, manage retries
Bidirectional communication: Report device status back to scheduler
Page accounting: Track pages actually printed for billing

Exit Codes:

Backends communicate status through exit codes:

0: Success
1: Retry later (temporary failure)
2: Abort job (permanent failure)
3: Authentication required
4: Job held for user action

Complete Job Processing Pipeline

Let's trace a complete print job from application to paper:

Step 1: Application Submission

Application calls CUPS API or writes to IPP socket
cupsd receives request, authenticates user
Job assigned unique ID, spooled to /var/spool/cups/

Step 2: Queue Management

Job added to appropriate printer queue
Scheduler evaluates priority and constraints
When printer ready, job selected for processing

Step 3: Filter Processing

Scheduler spawns filter chain as child processes
Each filter transforms data, writes to next stage
Final filter output is printer-ready format

Step 4: Backend Transmission

Backend receives filtered data
Establishes device connection
Transmits data, monitors for errors
Reports completion/failure to scheduler

Step 5: Cleanup

Spool files marked for deletion
Job history retained (configurable duration)
Notifications sent if requested

job_trace.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
#!/bin/bash
# Trace a print job through the CUPS system
 
# Submit test job, capture job ID
JOB_ID=$(lpr -P myprinter testfile.pdf 2>&1 | grep -oP 'request id is \K[^ ]+')
echo "Submitted job: $JOB_ID"
 
# Watch job progress
while true; do
    STATUS=$(lpstat -o $JOB_ID 2>/dev/null)
    if [ -z "$STATUS" ]; then
        echo "Job completed or not found"
        break
    fi
    echo "$STATUS"
    sleep 1
done
 
# Check spool files (requires root)
ls -la /var/spool/cups/d*-* 2>/dev/null
 
# View processing log
grep "$JOB_ID" /var/log/cups/page_log

Windows Print Spooler

Windows implements its own print spooler service (spoolsv.exe) with a different architecture but similar concepts.

Key Components:

Spooler Service: Core service managing queues and jobs
Print Processors: Transform EMF/RAW to printer format
Port Monitors: Handle device communication
Print Providers: Local vs network print routing

Spool File Formats:

EMF (Enhanced Metafile): Resolution-independent; rendered at print time
RAW: Already in printer's native format
TEXT: Plain text for basic output

Key Registry Locations:

HKLM\SYSTEM\CurrentControlSet\Control\Print\Printers
HKLM\SYSTEM\CurrentControlSet\Control\Print\Monitors

Spool Directory: C:\Windows\System32\spool\PRINTERS\

Print Spooler Security

The Windows Print Spooler has been the target of serious vulnerabilities (e.g., PrintNightmare CVE-2021-34527). It runs with SYSTEM privileges and handles remote connections, making it a high-value attack vector. Keep systems patched and consider disabling the service on systems that don't need printing.

Summary

Key Takeaways

•CUPS provides modern UNIX printing — Standards-based, uses IPP/HTTP throughout
•IPP is the modern printing protocol — Feature-rich, secure, bidirectional
•Filter chains transform documents — Automatic format conversion to printer-native
•Backends handle device communication — USB, network, virtual printers
•Complete pipeline from app to device — Well-defined stages with clear responsibilities

Next, we'll explore spool directories—the persistent storage that makes spooling reliable, including file organization, permissions, cleanup policies, and recovery mechanisms.

2 / 5

Loading learning content...

Operating SystemsI/O Software

Spooling: Simultaneous Peripheral Operations Online

LevelIntermediate

Duration60 mins

TopicI/O Software

2 / 5

Print Spooler: Architecture and Implementation

The Print Spooler: Where Theory Meets Practice

What You Will Learn

CUPS Architecture

Core Components:

cupsd - The scheduler daemon that manages print queues and jobs
cups-browsed - Discovers network printers and creates local queues
Filters - Convert document formats for specific printers
Backends - Handle communication with physical devices
PPD files - Describe printer capabilities and options

Design Philosophy:

CUPS uses HTTP/IPP (Internet Printing Protocol) for all communication, even locally. This unified approach means the same protocol works for local USB printers and remote network printers.

Converting Mermaid diagram...

cups_job_submission.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
/* CUPS Print Job Submission Example */
#include <cups/cups.h>
#include <stdio.h>
 
int submit_print_job(const char *filename, const char *printer) {
    cups_option_t *options = NULL;
    int num_options = 0;
    
    /* Configure print options */
    num_options = cupsAddOption("copies", "2", num_options, &options);
    num_options = cupsAddOption("media", "Letter", num_options, &options);
    num_options = cupsAddOption("sides", "two-sided-long-edge", 
                                 num_options, &options);
    num_options = cupsAddOption("print-quality", "5", num_options, &options);
    
    /* Submit job - returns immediately after spooling */
    int job_id = cupsPrintFile(printer, filename, "Document", 
                                num_options, options);
    
    cupsFreeOptions(num_options, options);
    
    if (job_id == 0) {
        fprintf(stderr, "Error: %s
", cupsLastErrorString());
        return -1;
    }
    
    printf("Job %d submitted to %s
", job_id, printer);
    return job_id;
}
 
/* Monitor job status */
void monitor_job(int job_id) {
    cups_job_t *jobs;
    int num_jobs = cupsGetJobs(&jobs, NULL, 0, CUPS_WHICHJOBS_ALL);
    
    for (int i = 0; i < num_jobs; i++) {
        if (jobs[i].id == job_id) {
            printf("Job %d: %s - %s
", 
                   jobs[i].id, jobs[i].title,
                   jobs[i].state == IPP_JSTATE_PENDING ? "Pending" :
                   jobs[i].state == IPP_JSTATE_PROCESSING ? "Processing" :
                   jobs[i].state == IPP_JSTATE_COMPLETED ? "Completed" :
                   "Unknown");
            break;
        }
    }
    cupsFreeJobs(num_jobs, jobs);
}

Internet Printing Protocol (IPP)

Key IPP Operations:

Operation	Purpose
Print-Job	Submit a new print job
Validate-Job	Check if job would be accepted
Get-Jobs	List jobs in a queue
Get-Job-Attributes	Query specific job status
Cancel-Job	Cancel a pending/processing job
Get-Printer-Attributes	Query printer capabilities
Hold-Job / Release-Job	Manage job scheduling

IPP Advantages:

Firewall-friendly: Uses standard HTTP/HTTPS ports
Secure: Supports TLS encryption and authentication
Feature-rich: Standardized attributes for all print options
Bidirectional: Clients can query status and receive events

IPP vs Legacy Protocols
Feature	IPP	LPD/LPR	Raw (Port 9100)
Transport	HTTP/HTTPS	TCP	TCP
Security	TLS, Authentication	None	None
Bidirectional	Yes	Limited	No
Job Status	Rich attributes	Basic	None
Standard	IETF RFC 8011	RFC 1179	De facto
Port	631	515	9100

The Filter Chain: Document Transformation

CUPS transforms documents through a filter chain—a pipeline of programs that convert the input format to the printer's native language.

Example Filter Chain:

PDF → pstops → pdftops → foomatic-rip → hp (backend)

Common Filters:

pdftops/pdftoraster: Convert PDF to PostScript or raster
pstops: Impose pages (n-up, booklet), select page ranges
foomatic-rip: Generic driver wrapper using PPD options
imagetopdf: Convert images to PDF for further processing

Filter Selection:

CUPS automatically determines the filter chain based on:

Input MIME type (detected or specified)
Printer's supported formats (from PPD)
Available filters in /usr/lib/cups/filter/

The scheduler traces a path through a MIME conversion graph to find the shortest chain from input to printer-native format.

simple_filter.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
/* Simple CUPS Filter Example
 * Filters read from stdin or file, write to stdout
 * Arguments: job-id user title copies options [filename]
 */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
 
int main(int argc, char *argv[]) {
    FILE *input;
    char buffer[8192];
    size_t bytes;
    
    /* Open input: file argument or stdin */
    if (argc > 6) {
        input = fopen(argv[6], "rb");
        if (!input) {
            fprintf(stderr, "ERROR: Cannot open %s
", argv[6]);
            return 1;
        }
    } else {
        input = stdin;
    }
    
    /* Log to stderr (appears in error_log) */
    fprintf(stderr, "DEBUG: Processing job %s for %s
", 
            argv[1], argv[2]);
    
    /* Process: this example just passes through */
    /* Real filter would transform the data */
    while ((bytes = fread(buffer, 1, sizeof(buffer), input)) > 0) {
        /* Apply transformation here */
        fwrite(buffer, 1, bytes, stdout);
    }
    
    if (argc > 6) fclose(input);
    return 0;
}

Debugging Filter Chains

Backends: Device Communication

Backends are the final stage—programs that transmit filtered job data to physical devices. Each backend handles a specific connection type.

Standard CUPS Backends:

CUPS Backend Types
Backend	URI Scheme	Purpose
usb	usb://make/model	USB-connected printers
socket	socket://host:9100	Raw TCP connection
ipp/ipps	ipp://host/path	IPP network printers
lpd	lpd://host/queue	Legacy LPD printers
smb	smb://host/share	Windows shared printers
serial	serial:/dev/ttyS0	Serial port printers
cups-pdf	cups-pdf:/	Virtual PDF printer

Backend Responsibilities:

Device discovery: Report available devices via backend --list
Connection management: Establish and maintain device connection
Error handling: Detect and report device errors, manage retries
Bidirectional communication: Report device status back to scheduler
Page accounting: Track pages actually printed for billing

Exit Codes:

Backends communicate status through exit codes:

0: Success
1: Retry later (temporary failure)
2: Abort job (permanent failure)
3: Authentication required
4: Job held for user action

Complete Job Processing Pipeline

Let's trace a complete print job from application to paper:

Step 1: Application Submission

Application calls CUPS API or writes to IPP socket
cupsd receives request, authenticates user
Job assigned unique ID, spooled to /var/spool/cups/

Step 2: Queue Management

Job added to appropriate printer queue
Scheduler evaluates priority and constraints
When printer ready, job selected for processing

Step 3: Filter Processing

Scheduler spawns filter chain as child processes
Each filter transforms data, writes to next stage
Final filter output is printer-ready format

Step 4: Backend Transmission

Backend receives filtered data
Establishes device connection
Transmits data, monitors for errors
Reports completion/failure to scheduler

Step 5: Cleanup

Spool files marked for deletion
Job history retained (configurable duration)
Notifications sent if requested

job_trace.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
#!/bin/bash
# Trace a print job through the CUPS system
 
# Submit test job, capture job ID
JOB_ID=$(lpr -P myprinter testfile.pdf 2>&1 | grep -oP 'request id is \K[^ ]+')
echo "Submitted job: $JOB_ID"
 
# Watch job progress
while true; do
    STATUS=$(lpstat -o $JOB_ID 2>/dev/null)
    if [ -z "$STATUS" ]; then
        echo "Job completed or not found"
        break
    fi
    echo "$STATUS"
    sleep 1
done
 
# Check spool files (requires root)
ls -la /var/spool/cups/d*-* 2>/dev/null
 
# View processing log
grep "$JOB_ID" /var/log/cups/page_log

Windows Print Spooler

Windows implements its own print spooler service (spoolsv.exe) with a different architecture but similar concepts.

Key Components:

Spooler Service: Core service managing queues and jobs
Print Processors: Transform EMF/RAW to printer format
Port Monitors: Handle device communication
Print Providers: Local vs network print routing

Spool File Formats:

EMF (Enhanced Metafile): Resolution-independent; rendered at print time
RAW: Already in printer's native format
TEXT: Plain text for basic output

Key Registry Locations:

HKLM\SYSTEM\CurrentControlSet\Control\Print\Printers
HKLM\SYSTEM\CurrentControlSet\Control\Print\Monitors

Spool Directory: C:\Windows\System32\spool\PRINTERS\

Print Spooler Security

Summary

Key Takeaways

•CUPS provides modern UNIX printing — Standards-based, uses IPP/HTTP throughout
•IPP is the modern printing protocol — Feature-rich, secure, bidirectional
•Filter chains transform documents — Automatic format conversion to printer-native
•Backends handle device communication — USB, network, virtual printers
•Complete pipeline from app to device — Well-defined stages with clear responsibilities

Next, we'll explore spool directories—the persistent storage that makes spooling reliable, including file organization, permissions, cleanup policies, and recovery mechanisms.

2 / 5

Spooling: Simultaneous Peripheral Operations Online - Page 2 of 5 | OneNoughtOne